Pa ae mace ra A eh a
Gn Spel shoal
ASS
Digitized by the Internet Archive in 2016 with funding from South African National Biodiversity Institute Libraries
https://archive.org/details/journalofsouthaf41 unse
RY GUNN LIBRARY
TMU ML
South Aeeneae ational Biodiversity Institute
JOURNAL
OF
SOUTH AFRICAN BOTANY
VOLUME 41 1975
Published under the authority of the
Trustees
of the
National Botanic Gardens of
South Africa, Kirstenbosch,
Newlands, C.P.
Editor: Prof. H. B. Rycroft
Director,
National Botanic Gardens of
South Africa,
Harold Pearson
Professor of Botany University
of
Cape Town
JOURNAL
OF
SOUTH AFRICAN BOTANY
VOLUME 41 1975 CONTENTS
NOTES ON THE DISTRIBUTION AND COM- PARATIVE LEAF MORPHOLOGY OF THE ACAULESCENT SPECIES OF S7RELITZIA AIT. H. A. van de Venter, J. G. C. Small and P. J. Robbertse ;
AN AUTORADIOGRAPHIC STUDY OF THE TRANSLOCATION OF !4C-LABELLED AS- SIMILATES IN ERAGROST/IS CURVULA (SCHRAD.) NEES AT DIFFERENT STAGES OF VEGETATIVE GROWTH. A. D. Barnabas and T. D. Steinke .
A QUALITATIVE STUDY OF THE NODULAT- ING ABILITY OF LEGUME SPECIES: LIST 3. N. Grobbelaar and B. Clarke .
ANDREAEA NITIDA HOOK. f. & WILS., A NEW RECORD OF A SOUTHERN HEMISPHERE MOSS FORAFRICA. E.A. C.L. E. Schelpe
A SHORT NOTE ON COMBRETUM EDWARD- S// EXELL. R. L. Verhoeven and H. P. van der Schijff : : ee er
BATCHELOROMYCES, A NEW GENUS FOR ANNELLIDIC DEMATIACEOQUS HYPHOMY- CETES ON PROTEA CYNAROIDES |N SOUTH AFRICA. W. F. O. Marasas, P. S. van ik and P. S. Knox-Davies :
PLANTAE NOVAE AFRICANAE: SERIES XXXIII. R. H. Compton .
BOOK REVIEWS: Flora of Lesotho by A. Jacot Guillarmod (E. A. Schelpe). 7rees of the Kruger National Park vol. 2 a P. van es e P. Rourke)
NOTES ON PROTEA IN SOUTH AFRICA: THE PROTEA TENAX TANGLE. J. P. Rourke
IDENTIFICATION OF ALLANTOIN IN PROTEA COMPACTA SEED. S. E. Drewes and J. van Staden F
STUDIES IN THE BULBOUS LILIACEAE IN SOUTH AFRICA: 5. J. P. Jessop
Page
17/
29
Sh7/
39
41
47
5S
63
67
DIE KARIOTIPE VAN OROTHAMNUS ZEYHERI (ER OnE CEES K. See and P. van der Merwe :
‘N NUWE NAAM VIR DUTHIELLA. M. P. de Vos
NEW TAXA AND NEW COMBINATIONS IN THE GENUS CRASSULA. H.R. Tolken
NOTES ON ERICA IN SOUTH AFRICA. H. A. Baker .
BOOK REVIEW: ere of South Africa BY C. Giddy (W. Wisura)
STUDIES IN THE GENUS CASS/A: 1. K. D. Gordon-Gray and D. P.K.Schorn . :
A NEW SPECIES OF PECTINARIA Oe Salee DACEAE). M. B. Bayer
STUDIES IN THE GENERA OF THE D/OS- MEAE (RUTACEAE) : 5.1. Williams
A SURVEY OF THE ANTIMYCOTIC AND ANTI- BACTERIAL ACTIVITY OF SOIL MICRO- FUNGI FROM TRANSVAAL. A. Eicker
EMBRIOGENIE VAN MELIANTHUS MAJOR L. E. Steyn
A NEW SPECIES OF PROTEA IN THE TRANS- VAAL. L.E. Davidson . :
EVIDENCE FOR THE PRESENCE OF A GERMI- NATION INHIBITOR IN SEEDS OF S7RELIT- ZIA A\T. H. A. van de Venter and J. G. C. Small . ;
INHIBITCRS ARE NOT RESPONSIBLE FOR POOR GERMINATION OF PROTEA COM- PACTA SEED. N. A. C. Brown and J. van Staden f ;
TERATOSPHAERIA PROTEAE-ARBOREAE AND MYCOSPHAERELLA JONKERSHOEKENSIS, TWO NEW ASCOMYCETES ON PROTEA IN SOUTH AFRICA. P. S. van Wyk, W. F. O. Marasas and P. S. Knox-Davies
STUDIES IN THE GENERA OF THE DIOSMEAE (RUTACEAE): 6. |. Williams .
NOTE ON THE MECHANISM OF POLLEN RELEASE IN BRUGUIERA GYMNORRHIZA. J. E. Davey
BOOK REVIEWS: Se caer of vascular plants by A. S. Foster and E. M. Gif- ford jr. (P. G. Jordaan). Fragile ecosystems edited by E. G. Farnworth and F. B. Golley (E. Moll). Aquatic plants of Australia 2) mle tle Ashton (J. P. Rourke) cae se re
INDEXTO PLANT NAMES : Vol. 41 (1—4)
87 91
93
125
131
188)
163
167
187
199)
207
211
225
231
239
269
273 275
THIS VOLUME IS DEDICATED TO
EDUARD MEINE VAN ZINDEREN BAKKER (1907— )
PHIL. NAT.D. (AMSTERDAM)
(Director of the Institute for Environmental Sciences, University of the Orange Free State, Bloemfontein)
Professor van Zinderen Bakker emigrated to South Africa in 1947 and became head of the Department of Botany of U.O.F.S. at Bloemfontein from 1963-1972. He initiated palynological studies in Africa and has become an authority on the history of the vegetation and climate of Africa during the Quaternary. Besides this he has achieved international recognition for his long term intensive bio- logical research programme on the sub-Antarctic islands Marion and Prince Edward. He also organised extensive limnological research.
Prof. Bakker still serves as chairman and member on four international research bodies in connection with Antarctic and palynological research. He is member of the editorial board of two international scientific journals. Among some of his rewards are the Senior Captain Scott Medal, the Gold Medal of the S.A. Ant- arctic Association, the Gold Medal of the S.A. Association of Botanists. He is a Fellow of the Royal Society of S.A., Officer in the Order of Orange Nassau and was Honorary Consul of the Netherlands (1953-1973).
JIS. Afr. Bot. 41 (1): 1-16 (1975)
NOTES ON THE DISTRIBUTION AND COMPARATIVE LEAF MOR- PHOLOGY OF THE ACAULESCENT SPECIES OF STRELITZIA ATT.
H. A. VAN DE VENTER, J. G. C. SMALL AND P. J. ROBBERTSE* (Department of Botany, University of Port Elizabeth)
ABSTRACT
Plants of S. reginae encountered in the wild represent a wide variety of leaf shapes and sizes. The junceous form of Strelitzia cannot be regarded as part of this variation as it is a distinct morphological entity found in extensive stands in the Uitenhage and Port Elizabeth areas. Seedlings of these two growth forms, which were raised in a growth cabinet, show morphologi- cal and ontogenetic differences right from the early growth stages. It is therefore proposed that S. juncea be regarded as a seperate species.
A population containing a variety of leaf forms intermediate between S. reginae and S. juncea was encountered near Patensie. The possibility that these plants are S. reginaex S. juncea hybrids needs to be investigated.
Although odd clones approximating the description of S. parvifolia do exist in the wild, the data presented in this paper suggest that this growth form does not merit specific rank.
UITTREKSEL
AANTEKENINGE OOR DIE VERSPREIDING EN VERGELYKENDE BLAARMOR- FOLOGIE VAN DIE ONGESTINGELDE SPESIES VAN STRELITZIA AIT.
Plante van S. reginae wat in die veld aangetref word, verteenwoordig ’n groot verskeiden- heid van blaargroottes en -vorms. Die biesieagtige groeivorm van Strelitzia kan nie as deel van hierdie variasie beskou word nie aangesien dit ’n eiesoortige morfologiese entiteit is wat in uitgebreide stande in die Uitenhage en Port Elizabeth omgewings voorkom. Saailinge van hierdie twee groeivorms, wat in ’n groeikabinet gekweek is, het vanaf die vroegste groeistadia verskil ten opsigte van hul morfologie en ontogenetiese ontwikkeling. Dit word dus voor- gestel dat S. Juncea as ’n afsonderlike spesie beskou word.
’n Bevolking wat naby Patensie aangetref is, bevat ’n verskeidenheid van blaarvorms intermediér tussen die van S. reginae en S. juncea. Die moontlikheid dat hierdie plante S. reginae x S. juncea basters is, regverdig verdere navorsing.
Alhoewel enkele klone wat aan die beskrywing van S. parvifolia voldoen, wel in die veld aangetref word, dui die gegewens wat in hierdie artikel aangebied word, daarop dat hierdie groeivorm nie as ’n afsonderlike spesie erken behoort te word nie.
INTRODUCTION
Two acaulescent species of Strelitzia are described in Flora Capensis (Wright, 1913) namely, Strelitzia reginae Ait. and S. parvifolia Ait.
Strelitzia reginae is described as having leaf blades approximately 45 cm (1,5 ft) long and 10 cm (4 in) wide with the margins undulate, especially in the lower part. Five varieties are differentiated on the basis of characteristics such as lamina. shape and size and petiole length. Four of these varieties are known from cultivated specimens only.
_* Present address: Department of General Botany, University of Pretoria. Accepted for publication 24th September, 1974.
1
2 Journal of South African Botany
Strelitzia parvifolia is distinguished from S. reginae on the basis of lamina size. The leaf blade of this species is reported to be approximately 23 cm (9 in) long and 7,5 cm (3 in) wide. The lamina has a flat margin. Only one variety (var. juncea Ker-Gawler) is listed and is described as a form which has either a very small leaf blade, or none at all.
Moore and Hyypio (1970) expressed the opinion that only one acaulescent species should be recognised, namely S. reginae. They suggest that as there are no discernible characters of the flower to separate S. parvifolia from S. reginae, and as this form probably represents no more than an anomalous clone of S. reginae, S. parvifolia does not merit specific rank.
Moore and Hyypio (1970) mention that the variant with no leaf blade was described independently in 1821 as S. juncea Link and S. parvifolia var. juncea Ker-Gawler, both from plants in cultivation. As distinctions from S. reginae other than in the lack of leaf blade are not apparent, and as its distribution in South Africa is not well documented, Moore and Hyypio decided to reduce S. juncea to S. reginae where, if desired, it could be designated as a cultivar— cy. ‘Juncea’.
Dr. R. A. Dyer of the Botanical Research Institute in Pretoria (personal communication) felt that this treatment was too drastic. In a publication on the vegetative multiplication of the acaulescent species of Strelitzia, Dyer (1972) recognizes two species, namely S. reginae Ait. and S. juncea Link.
In view of these conflicting opinions a study of the distribution and leaf morphology of S. reginae and its allies was undertaken.
MATERIALS AND METHODS
Various field trips were undertaken to examine Strelitzia species growing in the wild. Using data obtained in this way, as well as data made available by various herbaria, a distribution map of the acaulescent species was compiled. The opportunity was also used to plot the distribution of the caulescent species.
To permit morphological comparisons on a rational basis, measurement of various parameters (see Table 1) were taken on plants growing at three locali- ties. The following growth forms were compared in this way:
(a) Plants growing in Pluto’s Vale, 4 km from Ecca Pass (approximately
26 km from Grahamstown). Although a fair amount of variation in lamina shape and size was encountered in this population, the plants were all regarded as specimens of S. reginae because of the presence of a prominent leaf blade. None of the plants resembled the junceous form encountered in the Uitenhage area.
(b) Plants growing on the farm Brackhill, 26 km from Uitenhage on the
national road to Graaff-Reinet. These plants were very typical of the various uniform Strelitzia populations to be found in this area. The
Distr. and Compar. Leaf Morphology of Strelitzia Ait 3
leaves on all the plants either have extremely reduced laminae, or none at all. These plants are referred to as S. juncea in this text. None of the plants in this area were morphologically identical to any of the plants found at Pluto’s Vale.
(c) Plants growing in the Vensterhoek area about 16 km from Patensie. A variety of growth forms was encountered in this population. Some of the plants resembled the S. reginae specimens at Pluto’s Vale, others were similar to the junceous form found in the Uitenhage area, while a host of other plants was found with laminae of intermediate shapes and sizes.
The acaulescent species of Stre/itzia occur in the wild as isolated ‘‘clumps”’ (see Fig. 3). These specimens are not individual plants, as they superficially appear to be, but are a group of plants which have resulted from the vegetative multiplication of the original parent plant as described by Dyer (1972). The individual clumps encountered in the field will subsequently be referred to as clones.
Measurements were taken on ten randomly selected leaves and inflorescences in each clone. At Pluto’s Vale and Brackhill the survey was conducted on twenty randomly selected clones and mean values computed. Due to the striking variations in leaf morphology of plants growing at Venterhoek, the values recorded at this locality represent means from individual clones.
As environmental differences may result in morphological differences, a comparison of plants of S. juncea and S. reginae grown under identical con- ditions in a growth cabinet was also made. For this purpose seeds of S. reginae were collected from plants growing in the veld near East London while those of S. juncea were collected in the Uitenhage area. In both cases the plants were not growing near any differing growth forms so that the possibility that cross- pollination with such forms could have taken place was negligable.
The seeds were germinated aseptically in petri dishes and the seedlings transferred to a vermiculite substrate in plastic containers of one litre capacity. As growth proceded the plants were transplanted to larger pots. Nutrients were supplied by means of biweekly applications of a complete nutrient solu- tion. The relative proportions of the different macro-elements in the solution were based on the formulation used by Murray (1959) for Musa species.
The plants were grown in a Percival model PGW-132 growth cabinet at day/night temperatures of 26/18 °C. A photoperiod of 14 hours with a light intensity of approximately 20 000 lux was maintained.
Leaf measurement of the plants were taken from time to time.
RESULTS The distribution of Strelitzia species in South Africa is depicted in Fig. 1.
4 Journal of South African Botany
8 Louis Trichardt
Pietersburg
S.caudata { TRANSVAAL
S. nicolai
4
y) t
ea ;
/
7 ORANGE FREE STATE
CAPE PROVINCE
Fic. 1. Distribution of Strelitzia species in South Africa.
The herbarium specimens used in the compilation of this map are listed in the appendix.
S. reginae
This species is distributed along the east coast of South Africa from Patensie to Zululand. It could not be ascertained whether the plant has ever been found in the wild on the south coast of Natal.
The largest populations of this species were found at Ecca Pass and Pluto’s Vale, near Grahamstown. The average size of laminae in the Pluto’s Vale population was found to be smaller than the size recorded in Flora Capensis for the species. In Table 1 it can be seen that the mean lamina length was 33 cm and mean lamina width 10 cm. However, a large amount of variation in leaf size was found. A clone with laminae on the average 41 cm long and 5 cm wide was encountered as well as one with laminae 29 cm long and 8,5 cm wide.
Distr. and Compar. Leaf Morphology of Strelitzia Ait 5
In Fig. 2 an extreme example is pictured—teaf blades 30 cm long and 3 cm wide.
Fic. 2. Leaves and flowers of a S. reginae clone found at Pluto’s Vale.
Considerable variation in lamina size, as well as lamina shape, was also encountered among S. reginae plants growing elsewhere. It would seem that leaf size is influenced to a large degree by environmental conditions. Plants in the wild growing in dense bush or a clump of trees invariably have abnormally long petioles—and yet this is a criterion which is used to differentiate between varieties in Flora Capensis. Plants in cultivation often grow to a greater height than those found in the wild.
Botany
Journal of South African
“OOAQ[OS APYSBI[S S9OIde JeIT xx» ‘OOIOTOS A[[VWIOU SsdIdeV JeaT x».
“SQUO]D [VNPIAIPUL JO SJUDWAINSRSLU JUBSO.IC91 YOOYIO}SUSA IOJ SONA “SqUO]D YZ WO SosVIOAL JUOSoIdOI [IY OeIg PUL IIA S.OIN]d IOJ SONJA x
Gl! iL, iG € 8 (0) {L 9 oa | Ali “9uo[o Jod SIDUDISIIOLYUL JO JToquunny 07 81 81 07 61 = IZ v7 07 07 a eee tae er OCO SIE S 96 €8 80. Ol v6 — L6 LOT vor 88 0 8 9 5 5 9 8 WERE COEDS x0) I ST G € 7 50 II Ol pI ATU «x9T eS ex TT xxx 07 QE EL aS 6@ ek €€ sot yysuoy BUTE T esi | OCCA (GLC Ce OE! 86 Sol SL so ysua] BTONAd 691 Srl 9€L vel OT LET SEI Lol Sol 801 ss Feat JO YISUI] [eIOL Ofl 00% OCI SL Orr 00r os OSI 06 EST + ({eAg] puNoIs) JoJQWUIeIP 9UO[D, O6l OLT OLT O9f 00¢ OLT 091 al Sol Gal (Jea] }SeSUOT JO YISUIT) 1YSI9Y4 UO]
SUIIOJ O}VIPSUUIO}UT paounl *s apuisad*§ vaount{ ss apursas “S soisadg
yoouso}susA, [Tupoesg, [vA AypesoT
$.01nT
DIZINAAG JO SUIIOJ YIMOIZ yUDSSETNVOe JO ,SuOHe[Ndod sory} Jo (WIS UI) SUOTsUSUIP StUOS—] ATA VL
Distr. and Compar. Leaf Morphology of Strelitzia Ait 7
In the light of the extensive variation in leaf shape and size of plants of S. reginae, two points may be emphasized :
(a) It is meaningless to delineate the species in terms of leaf shape and size unless very wide limits are set.
(b) The indescriminate recognition of varieties on the basis of criteria such as these, especially from plants in cultivation, can serve no useful purpose and only lead to confusion.
S. juncea
An acaulescent form of Strelitzia which is quite distinct from S. reginae as far as appearance is concerned, occurs in relatively extensive stands at different localities north west of Uitenhage. Two populations are also located in the Port Elizabeth area; one at Coega Kop and the other in Gelanse Kloof. Jun- ceous forms are also found at Vensterhoek.
Plants of this species have a bizarre appearance (see Fig. 3). The foliage consists of long, erect petioles which, as a rule, have sclerotic apices (d in Fig. 4). In some cases the petiole tip dies back for a considerable distance (e in Fig. 4).
When viewed from a distance, all the leaves of this growth form seem to be devoid of laminae. On closer examination, however, it can be seen that the
Fic. 3. Clones of S. juncea in the veld near Uitenhage.
8 Journal of South African Botany
181
Fic. 4. Leaf-forms encountered on S. juncea plants.
petioles of some of the plants support a very small lamina which is approx- imately 4 cm long and 0,5 cm wide (b in Fig. 4). Some leaves do not have small laminae or sclerotic tips, but only a small groove in the petiole apex (c in Fig. 4).
Leaves with laminae measuring approximately 10 cm by 2-4 cm are also encountered on S. juncea plants (a in Fig. 4). Such laminae are occasionally found on mature plants and are carried on very short petioles. They are commonly seen, however, as regrowth after the foliage of mature plants has been cut down to almost ground level. As seedlings of S. juncea produce leaves
Distr. and Compar. Leaf Morphology of Strelitzia Ait 9
which are distinctly laminate (see Fig. 5), the large lamina probably represents a juvenile leaf form.
The data in Table 1 show that the S. juncea plants examined were taller than the plants of S. reginae. Although these figures were obtained from plants growing at different localities and are, therefore, not strictly comparable, they probably do reflect genetic differences. Where the two species are cultivated in close proximity, such as at Kirstenbosch and in various parks in Port Elizabeth, it can be seen that the plants of S. juncea are invariably taller than those of S. reginae.
The leaves of plants of S. juncea at Brackhill were almost 60 cm longer than leaves of S. reginae plants in Pluto’s Vale (see Table 1). It is also important to note that the average leaf length of 165 cm in the case of S. juncea actually represents the length of the petiole only. In the case of S. reginae, the average leaf length of 108 cm includes 75 cm of lamina. The petioles of these plants are therefore less than half as long as the petioles of S. juncea.
Although measurements were not taken of any of the plants growing in the Port Elizabeth area, these plants seem to resemble those growing near Uitenhage in all respects.
Growth forms at Vensterhoek
A very interesting population of acaulescent Stre/itzia plants was encountered at Vensterhoek near Patensie. In this rather scattered population are plants which resemble S. reginae as well as plants which resemble S. juncea. The reduced laminae of the junceous plants are, however, slightly longer than is the case for plants of this form in the Uitenhage area. The junceous forms at Vensterhoek frequently have leaves which have died back from the tip so that only the sclerotic remains of these laminae are seen.
Also to be found in this population are plants representing a large variety of lamina shapes and sizes. Some of these plants are described by the data in Table 1 and it can be seen that some clones have laminae of unusual pro- portions in comparison with laminae of “typical” S. reginae and S. juncea plants.
The variation found at Vensterhoek is reminiscent of the so-called “hybrid scatter’ which occurs after cross-pollination between species. It is possible that these intermediate forms are S. reginae x S. juncea hybrids. This possibility will be investigated in the near future.
S. parvifolia
As far as could be ascertained, there are no herbarium specimens of this species in South Africa. The only record of the species which could be traced is a photograph of a specimen in the herbarium of the Royal Botanic Gardens
10 Journal of South African Botany
at Kew. The lamina of this specimen is approximately 22 cm long and 4 cm wide. The only clue as to locality is the annotation “Cape”. It is interesting that Flora Capensis does not mention any specific locality for this species either.
No uniform stand of plants conforming to the description of S. parvifolia or resembling the specimen at Kew were encountered anywhere in the veld. There are, no doubt, clones at Pluto’s Vale and Vensterhoek which would fit the description of S. parvifolia. However, in the light of the variation between the different clones it would be ridiculous to accept any such variant as a separate species.
No other conclusion can be drawn but that S. parvifolia does not merit specific rank. This is in agreement with the opinion of Moore and Hyypio (1970).
Comparative morphology of seedlings in a controlled environment
No differences in external morphology are evident as far as germinating seeds of S. reginae and S. juncea are concerned. Differences are, however, apparent in the seedlings which develop from these seeds.
The laminae of S. reginae seedlings are larger than those of S. juncea, a difference which can be seen even after the lamina of the first laminate leaf has extended (see Fig. 5a).
As successive leaves of S. reginae seedlings appear, a progressive increase in lamina size is found. A different pattern of development is encountered in seedlings of S. juncea. (See Figs. 5 and 6).
Laminae of S. juncea seedlings show an increase in size only up to about the eighth leaf, depending on the individual seedling. Hereafter successive leaves show a progressive decline in lamina size. The lamina of the twelfth leaf of the S. juncea specimen shown in Fig. 5c (that is, the longest leaf shown in the figure) is barely visible and is typical of laminae found on mature plants of S. juncea. This lamina was found to be 4,0 cm long and 0,7 cm wide.
The first few leaves of S. reginae have longer petioles than the corresponding leaves on S. juncea seedlings. From the eighth leaf onwards, however, S. juncea produces longer petioles than S. reginae.
All the leaves of S. juncea have a higher ratio of petiole length to lamina length in comparison with leaves of S. reginae. The tenth leaf of S. juncea plants, for instance, has a petiole which is, on the average 4,89 times longer than the lamina. The petiole of the tenth leaf of S. reginae plants is only 1,38 times longer than the lamina. The twelfth leaf of the S. juncea specimen pictured in Fig. 5 has a petiole length: lamina length ratio of 20.
Differences in lamina shape between the two species were also recorded. The laminae of S. juncea are generally elliptic. The first few leaves of S. reginae also have an elliptic form while those appearing later are decidedly lanceolate.
11
Distr. and Compar. Leaf Morphology of Strelitzia Ait
12 Journal of South African Botany
Fic. 5.
Comparison of growth of seedlings of S. juncea (no. 6) and S. reginae (no. 13a) in a controlled environment a. Germinating seeds; b. Plants after eight months of growth; c. Plants after 15,5 months of growth.
DISCUSSION
The data presented in this paper confirm the view of Moore and Hyypio (1970) that S. parvifolia does not merit recognition at specific level. Plants of S. reginae encountered in the wild exhibit a wide range of leaf shapes and sizes and some of these clones probably approximate the description of S. parvifolia as recorded in Flora Capensis.
The junceous form cannot be regarded as part of the variation within S. reginae. It is a distinct morphological entity and can be distinguished from
Lamina width (cm)
Distr. and Compar. Leaf Morphology of Strelitzia Ait 13
[|
L
of imal 3b | ] ef | [) Ai 6r
=> L 5 |
2
4P
Sp 1 IL
2}
3 27
&
a [ l|
LI &§ LI Wu lI Lu 12346465 678 9 Oh P 123 €6 678 9 M Tl ® Leaf number Leaf number a. b. Fic. 6.
Lamina measurements of a young S. juncea plant a. Lamina width b. Lamina length.
S. reginae under a wide range of environmental conditions. The ontogenetic development of the plants also differs from the early growth stages of the seedlings.
The distribution of S. juncea is also distinct from that of S. reginae. Apart from plants in cultivation, a mixed stand of the two species was only found at one locality, namely Vensterhoek. Admittedly the various growth forms found at this locality do present a classification problem and further research is required on this population. The possibility that these intermediate growth forms could be S. reginae x S. juncea hybrids needs to be investigated. Apart from cross-pollination studies, the approach of Frést and Bose (1966) and Frost and Ising (1968) using chemical systematic methods might prove valuable in this respect.
Attempts at a cytological comparison of S. reginae and S. juncea have been made but the mitotic figures obtained so far have not been sufficiently clear to justify a definite conclusion.
In the light of the data presented in this paper it is suggested that the junceous form maintains the name S. juncea Link. Furthermore, it is proposed that the key presented by Moore and Hyypio (1970) be modified with respect to the acaulescent species of Strelitzia. The following is proposed:
Flowers with orange to yellow sepals and deep blue-violet petals; penduncles of inflorescences elongate, erect from the leaf axils; erect stems not developed.
Leaf blades prominent, at least 3 cm wide and 30 cm long, usually wider and longer (1) S. reginae
14 Journal of South African Botany
Leaf blades absent, or very reduced. Petioles usually terminate in dead apex. (2) S. juncea
In the case of S. reginae such a diversity of lamina shapes and sizes is en- countered that it would hardly be rewarding to attempt a classification at the intraspecific level.
ACKNOWLEDGEMENTS Financial assistance from the Council for Scientific and Industrial Research as well as the University of Port Elizabeth, is acknowledged.
REFERENCES Dyer, R. A., 1972. Vegetative multiplication of Strelitzia reginae and its allies. Bothalia 10: 57 8.
Frost, S. AND Bose, S., 1966. An investigation of phenolic compounds in two species of jute (Corchorus olitorius and C. capsularis) and their supposed hybrid, using thin layer chromatographic technique. Hereditas 55: 179-187.
Frost, S. AND ISING, G., 1968. An investigation into phenolic compounds in Vaccinium myrtillis L. (Bilberries), Vaccinium vitis idea L. (Cowberries) and the hybrid between them V. intermedium Ruthe employing thin layer chromatography. Hereditas 60: 72-76.
Moore, H. E. NR. AND Hyypio, P. A., 1970. Some comments on Stfre/itzia (Strelitziaceae). Baileya 17: 65-74.
Murray, D. B., 1959. Deficiency symptoms of the major elements in the banana. Trop. Agric., Trin. 36: 100.
WriGHT, C. H., 1913. Scitaminiae. In: Flora Capensis, vol. V, section 3, edited by W. T. Thiselton-Dyer. London: L. Reeve and Co.
APPENDIX
The following herbarium specimens were used for the compilation of the distribution map of Stre/itzia (Fig. 1):
S. alba
CAPE PROVINCE—3323 (Willowmore): Groot Rivier Pass (-CD), abundant on steep slopes and occasional in forest near streams, Dyer 5329 (PRE); on eastern slopes, Small and Robbertse 3 (PEU); in forest and on ledges of cliffs, Keet s.n. (GRA); Blauwkrantz (-DC), Groot Kloof, Schmidt s.n. (GRA); Storms River (-DD), on steep slopes of thickly wooded kloof, Fourcade s.n. (GRA).
S. caudata
TRANSVAAL—2329 (Pietersburg): Pietersburg (-DD), Woodbush Forest Station, Grewcock s.n. (PRE). Type: 2330 (Tzaneen): Zoutpansberg (-AA), from farm “Geluk’”’, 23 km East of Louis Trichardt, Veschuur s.n. (PRE); Piesanghoek (-AA), amongst rocks in dense bush on summit and on rocks on bush-covered krans on mountainside, Galpin s.n. (PRE); Louis Trichardt (-AA), in Widdringtonia bush of Piesanghoek, Gerstner 5784 (PRE); West- phalia (-CC), Pietersburg district, Burtt-Davy s.n. (PRE); Letaba (-CD),
Distr. and Compar. Leaf Morphology of Strelitzia Ait 15
Scheepers 428 (PRE). 2430 (Pilgrim’s Rest): Pilgrim’s Rest (-DD), v.d. Schijff 6462 (PRE). 2531 (Komatipoort): Barberton (-CC), Dyer 4802 (PRE). SWAZILAND—2531 (Komatipoort): Pigg’s Peak (-CD), Heenan s.n. (PRE).
S. juncea
CAPE PROVINCE—3324 (Steytlerville): Vensterhoek (-DD), on farm Vensterhoek, Small and Robbertse 28 (PEU); van de Venter 147 (PEU). 3325 (Port Elizabeth): Brackhill (-CB), on northern slope, Small and Robbertse 6; 55 (PEU); Prentice Kraal (-CB), on northern slopes on side of Uitenhage— Jansenville road, van de Venter 181 (PEU); Uitenhage (-CB), scrubveld 5 km from Uitenhage on Steytlerville road, Admiraal 468 (PRE); 16 km from Uiten- hage on Steytlerville road in Karroid scrub, Dyer 4832 (PRE); Dyer and Wells 5820 (PRE); on mountains, Patterson 665 (GRA); Uitenhage, Thode A 2752 (PRE); Port Elizabeth (-DA), on northern slopes in Gelanse Kloof, above Salsoniville, van de Venter 213 (PEU); Bethelsdorp, Patterson 665 (GRA).
S. nicolai
NATAL—2632 (Bela Vista): Zululand (-DD), Kosi Bay, Rodni 4610 (PRE). 2732 (Ubombo): Ubombo (-CA), 12 km S.W. of Manzengwenya on road to Mbazwana, de Winter and Vahrmeijer 8589 (PRE). 2831 (Nkandla): Hlabisa (-BB), Ward 4417 (PRE). 2832 (Mtubatuba): Mtubatuba (-AC), coastal sand- flats, scrub with scattered trees, Palm Ridge farm, Harrison 414 (PRE). 2930 (Pietermaritzburg): Isipingobeach (-DD), Ward 854 (PRE). 3030 (Port Shep- stone): Margate (-CD), Prosser s.n. (PRE); Southport (-DA), Bruce 437 (PRE).
CAPE PROVINCE—3228 (Butterworth): (-CC), on bank of lagoon, van de Venter 60 (PEU).
S. reginae
NATAL—2831 (Nkandla): Nxeba Ridge (-BD), hill near junction of Black and White Umfolozi Rivers, Barret s.n. (PRE); Mpila Hill, Umfolozi Game Reserve (-BD), on steep slope amongst rocks, Borquin 533 (PRE); Heatonville (-DB), on banks of spruit, Houting s.n. (PRE); Empangeni (-DD), in bush on banks of stream, Houting s.n. (GRA).
CAPE PROVINCE—3129 (Port St. Johns): Mtata River (-CC), short distance north of Coffee Bay on coast, Vogts s.n. (PRE). 3227 (Stutterheim): near East London—Nahoon road (-DD), van de Venter 169 (PEU); Arnoldton (-DD) on farm “‘Fairview’’, van de Venter 176 (PEU). 3228 (Butterworth): Kei Mouth (-CB), Flanagan 1706 (PRE); 6 km from river mouth, Small and Rob- bertse 21 (PEU); Cintsa (-CC), 5 km from coast on banks of stream, Small and Robbertse 16 (PEU); Kwelera (-CC), on stream bank in shade of gully bush, 75 km from Kwelera on road to East London, Comins s.n. (GRA). 3324 (Steyt- lerville): Vensterhoek (-DD), on farm Vensterhoek, Small and Robbertse 32;
16 Journal of South African Botany
33; 150 (PEU); van de Venter 149 (PEU). 3326 (Grahamstown): Ecca Pass (-BA), eastern slopes, Small and Robbertse 8; 34 (PEU); Pluto’s Vale (-BA), Small and Robbertse 56; 57; 58; 170-174 (PEU); Kariege Rivier (-DA), im walde Zeyher 3953 (GRA); Port Alfred (-DB), on banks of stream under dense bush, Small and Robbertse 10 (PEU). 3327 (Peddie): Keiskamma River (-AB), 16 km from mouth in dense bush, Small and Robbertse 11 (PEU); East London (-BB), hillsides along Buffalo River, Galpin s.n. (PRE); Keiskamma (-BC), Sim 19927 (PRE).
JIS. Afr. Bot. 41 (1): 17-27 (1975)
AN AUTORADIOGRAPHIC STUDY OF THE TRANSLOCATION OF 4C-LABELLED ASSIMILATES IN ERAGROSTIS CURVULA (SCHRAD.) NEES AT DIFFERENT STAGES OF VEGETATIVE GROWTH
A. D. BARNABAS AND T. D. STEINKE (Department of Botany, University of Durban-Westville)
ABSTRACT
Autoradiography was used to study the movement of !4C-labelled assimilates in Eragrostis curvula at several stages of vegetative growth. The assimilates were labelled by allowing single leaves of plants to photosynthesize in an atmosphere containing radioactive carbon-dioxide (4CO,). Leaves of different ages, located on different tillers, were treated.
Assimilates were exported mainly to growing regions, such as young developing leaves, and to the roots. The pattern of distribution of assimilates changed during the growth of the plant. During the early stages of development (seedling to four-tiller stage) there was extensive intertiller translocation between the main and daughter tillers and vice versa. During the later stages of growth (six-tiller to mature vegetative stage) intertiller translocation decreased considerably and tillers became more independent.
UITTREKSEL
’N OUTORADIOGRAFIESE STUDIE VAN DIE TRANSLOKASIE VAN C-GE- MERKTE ASSIMILATE IN ERAGROSTIS CURVULA (SCHRAD.) NEES BY VER- SKILLENDE VEGETATIEWE GROEISTADIA.
Outoradiografie is gebruik om die beweging van '*C-gemerkte assimilate in Eragrostis cur- vula by verskillende vegetatiewe groeistadia te bepaal. Assimilate is gemerk deur enkele blare van plante te laat fotositetiseer in ’n atmosfeer wat radio-aktiewe koolstofdioksied (44CO,) bevat. Blare van verskillende ouderdomme, en op verskillende lote is so behandel.
Assimilate is hoofsaaklik vervoer na groeiende dele soos jong, ontwikkelende blare, en na die wortels. Die distribusiepatroon van assimilate het verander terwyl die plant gegroei het. Gedurende die vroeé groeistadia (kiemplant tot 4-loot stadium) was daar uitgebreide interloot- translokasie tussen die hoof- en dogter-lote en omgekeerd. Gedurende latere groeistadia (6-loot tot volwasse vegetatiewe stadium) het interloot-translokasie aansienlik afgeneem, en het lote meer onafhanklik geword.
INTRODUCTION
The translocation of !*C-labelled assimilates in members of the Gramineae has been the subject of several investigations. Well-defined distribution patterns of assimilates have been found in cereals, such as wheat (Quinlan and Sagar, 1962; Lupton, 1966) and in grasses such as Phleum pratense L (Williams, 1964) and Lolium multiflorum Lam. (Sagar and Marshall, 1966; Marshall, 1967). These studies have shown how the various parts of a graminaceous plant are interrelated in terms of assimilate movement. The object of the present study was to contribute to work done in this direction since little is known about the distribution patterns of photosynthetic assimilates in grasses indigenous to
Accepted for publication 18th October, 1974. 9/
18 Journal of South African Botany
South Africa. Eragrostis curyula was chosen because it is not only a common veld grass but is also important agronomically.
MATERIAL AND METHODS
Cultural conditions:
Seeds of Eragrostis curyula (Ermelo cultivar) were sown in 15 cm clay pots containing a 1:1 mixture of sand and soil. The sand and soil were autoclaved at a pressure of | bar for 15 min prior to filling the pots. The plants were grown in a temperature-controlled glasshouse which was provided with two air- conditioning units. This ensured that the plants were not exposed to very high or low temperatures. Natural daylight was extended by six 200 watt mercury- tungsten lamps to give a 16h photoperiod. Watering was carried out as fre- quently as necessary to prevent any moisture shortage. Small applications of a commercial fertilizer were made uniformly to each pot.
Under these conditions plants were healthy and grew vigorously. Seedlings were selected for uniformity and thinned to four per pot. When most plants had two tillers, selection for uniform development was again carried out, and the plants were thinned to one per pot.
Assimilation of “CO,
The method used to introduce labelled carbon into plants was based on a technique similar to that of Quinlan and Sagar (1962). Single leaves of plants were allowed to photosynthesize in an atmosphere containing “CO,. Radio- active carbon-dioxide was produced by the addition of excess 50 per cent lactic acid to an aqueous solution of “C-labelled sodium carbonate (Na,!*CO;). The specific activity of the sodium carbonate was 10 mc/mM.
In preliminary experiments it was found that a good autoradiographic image could be produced if small plants (seedling to six-tiller stage) received a dose of 0,5 ue of radioactivity and bigger plants (eight-tiller to mature vegeta- tive stage) 1 wc. A larger dose was required for larger plants because of dilution of the tracer within the plant.
The administration of ‘CO, was performed in the glasshouse in bright sunlight on a clear day. The assimilation period lasted for 6h, from 0930 to 1530h. The plants were left in the glasshouse and harvested 18h later.
Autoradiography:
Soil was carefully washed free from the roots and the plants were freeze- dried and autoradiographed following a technique modified from Yamaguchi and Crafts (1958), Little (1962) and Crafts and Yamaguchi (1964).
Mounted plants were exposed to Kodak Blue Brand X-ray film for two weeks if the radioactive dose was 0,5 yc or for 1 week if 1 pc of radio-activity
Translocation of Assimilates in Eragrostis Curyula 19
was given. The manufacturer’s instructions for processing X-ray films were followed.
Treatments:
Treatments were carried out at seven different stages spanning the develop- ment of the plant. These stages together with the leaves which were treated are shown in Table 1.
TABLE 1.
Leaves which assimilated Co, in plants at different stages of development.
STAGE OF DEVELOPMENT TREATMENT LEAVES (i) Seedling (ii) 1-Tiller — Youngest (iii) 2-Tiller fully Youngest fully expanded Youngest | expanded leaf (iv) 4-Tiller leaf on fully on youngest main expanded | daughter tiller (v) 6-Tiller tiller tiller leafon = |————_—_—— oldest — (vi) 8-Tiller daughter
tiller as above
(vii) Mature Vegetative
In this investigation, the primary or dominant tiller was referred to as the main tiller and the secondary or subsidiary tillers were called daughter tillers. A diagrammatic representation of a plant at the four-tiller stage is illustrated in Fig. 1. The main tiller (MT) is shown, together with the four daughter tillers (T;, Tz, T3, T,) at different stages of development (T,, oldest and T, youngest). The youngest fully expanded leaf (indicated by numeral I) on the main tiller and on the oldest and youngest daughter tillers, is shown. A fully expanded leaf was recognized by a clearly visible ligule. In addition, the oldest green leaf on the main tiller (II) is shown.
Each plant was treated on a single leaf, different plants being used to deter- mine the patterns of distribution of labelled assimilates from the different leaves. Uniform plants were selected for treatment. Each treatment was repli- cated twice.
RESULTS
The results are presented as visual assessments of autoradiographs. Photo- graphs of mounted plants and their corresponding autoradiographs are shown. Since there was very close agreement between replicates in the majority of
20 Journal of South African Botany
Wa j ae | Wea Ye i WW YF I Via JZ ff ] \ \ \ 7 aah NN Fic. 1.
Diagrammatic representation of a plant at the four-tiller stage of development.
treatments, only one replicate set, unless otherwise stated, is presented. In the photographs of the mounted plants, the treated leaf is indicated by an arrow.
The patterns of distribution of labelled assimilates during the various vegetative stages of growth, following assimilation of *CO, by single leaves of different ages located on either the main or daughter tillers are shown in Figs. 2-13.
(a) Treatment of youngest fully expanded leaf on main tiller:
In the seedling stage (Fig. 2, plant at upper left hand corner), 'C-assimilates from the treated leaf moved mainly to those parts which were actively growing, accumulating in the young developing leaf, in the lower part of the expanding leaf, in the roots and to a lesser extent in the mature leaf.
An unlabelled plant, at the lower right hand corner of Fig. 2, was treated in exactly the same way as the other three labelled plants. As can be seen, it did
Translocation of Assimilates in Eragrostis Curvula 21
~ / oN ; | &
2 3 4
Fics. 2-4
Photographs of mounted plants above, autoradiographs below. Fig. 2. Treatments of plants
at the seedling stage: Youngest fully expanded leaf treated (plant at upper left hand corner);
oldest green leaf treated (plants at upper right and lower left hand corners); and an untreated
plant (lower right hand corner). Fig. 3. Youngest fully expanded leaf on main tiller of a dupli-
cate set of plants at the one-tiller stage treated. Fig. 4. Youngest fully expanded leaf on main tiller of a plant at the two-tiller stage treated.
not reveal any significant autoradiographic artefacts which could have in- fluenced the interpretation of the other autoradiographs.
The labelling pattern of plants at the one, two and four-tiller stages (Figs. 3, 4 and 5 respectively) were similar. Carbon-14 assimilates were exported mainly to young developing leaves of the main and daughter tillers and to the roots. In almost all other leaves there was a faint trace of radiocarbon.
A change in the distribution pattern of assimilates occurred at the six-tiller stage of development (Fig. 6). Export of 'C-assimilates to the daughter tillers decreased considerably, although the roots continued to accumulate radio- carbon.
At the eight-tiller stage there was a similar pattern of distribution of “C- assimilates as at the six-tiller stage.
Export of labelled assimilates from the treated leaf decreased even more at the mature vegetative stage (Fig. 7). Only the roots received assimilates,
22 Journal of South African Botany
Kx
p aN i . a Hf eget ea Ee en Ve
ee f See.
5 6 7
Fics. 5-7
Photographs of mounted plants above, autoradiographs below. Youngest fully expanded leaf on main tiller of plants at the four-tiller stage (Fig. 5), six-tiller stage (Fig. 6) and mature vegetative stage (Fig. 7), treated.
(b) Treatment of oldest green leaf on main tiller:
The patterns of distribution of labelled assimilates from the oldest green leaf on the main tiller, were very similar to those from the youngest fully expanded leaf on the main tiller at all vegetative stages of growth, in which treatments were imposed. Only the results for the seedling stage (Fig. 2) and for the four and six-tiller stages (Figs. 8 and 9 respectively), are shown.
Again, “C-assimilates from the treated leaf were translocated predominantly to young developing leaves of the main and daughter tillers and to the roots in plants up to the four-tiller stage of development.
From the six-tiller stage to the mature vegetative stage export of assimilates to the daughter tillers again decreased considerably although the roots con- tinued to accumulate radiocarbon.
(c) Treatment of youngest fully expanded leaf on oldest daughter tiller:
At the two and four-tiller stages labelled material was translocated mainly to young developing leaves of the treated tillers, to the root systems and to a
Translocation of Assimilates in Eragrostis Curvula 0B
I ES Ay aN eS EGP
8 9 10
Fics. 8-10
Photographs of mounted plants above, autoradiographs below. Oldest green leaf on main tiller of plants at the four-tiller stage (Fig. 8) and six-tiller stage (Fig. 9), treated. Youngest fully expanded leaf on oldest daughter tiller of a plant at the four-tiller stage (Fig. 10), treated.
lesser extent to growing regions of the main and other daughter tillers. Traces of radioactivity were present in most mature leaves. An autoradiograph of a plant at the four-tiller stage (Fig. 10) only is shown.
Again a change in the distribution pattern of assimilates occurred from the six-tiller stage to the mature vegetative stage. Carbon-14 assimilates from the treated leaves of plants at the six- and eight-tiller stages and the mature vegeta- tive stage, were translocated mainly to growing regions of the treated tillers, to developing leaves of young tillers nearest them and to the roots. It is interesting to note that no assimilates were exported to the main tillers and translocation was restricted only to tillers on the same side as the treated tillers. An auto- radiograph of a plant at the six-tiller stage (Fig. 11) only is shown.
(d) Treatment of youngest fully expanded leaf on youngest daughter tiller:
The distribution of labelled assimilates at the two and four-tiller stages, was similar to the distribution patterns that were obtained in plants at the same developmental stages, when the youngest fully expanded leaf on the oldest
24 Journal of South African Botany
Fics. 11-13 Photographs of mounted plants above, autoradiographs below. Youngest fully expanded leaf on oldest daughter tiller of a plant at the six-tiller stage (Fig. 11), treated. Youngest fully expanded leaf on youngest daughter tiller of plants at the four-tiller stage (Fig. 12) and eight- tiller stage (Fig. 13), treated.
daughter tiller, was treated. An autoradiograph of a plant at the four-tiller stage (Fig. 12) is shown.
In a plant at the eight-tiller stage of development (Fig. 13) almost all the assimilated carbon was retained by the treated leaf. Some labelled carbon was exported to young developing leaves of the treated tiller and to the roots.
DISCUSSION
The patterns of distribution of labelled assimilates in Eragrostis curvula agree generally with findings in other members of the Gramineae in which similar investigations were carried out (Quinlan and Sagar, 1962; Williams, 1964; Sagar and Marshall, 1966; Marshall, 1967). All growing regions which are incapable of sustaining their own carbon requirements for rapid growth, make a demand upon the flow of assimilates from mature leaves. Assimilates, however, do not go indiscriminately to all growing regions. The stage of develop- ment and the position and age of the leaf that fixed “C, influenced the distribu- tion patterns.
Translocation of Assimilates in Eragrostis Curvula DS
During the early stages of growth (the two and four-tiller stages) there was extensive movement of assimilates from the main tiller to the daughter tillers and vice versa. This reciprocal movement has also been reported by other workers. In wheat plants at the two-tiller stage, Quinlan and Sagar (1962), found that the main shoot passed labelled material to the two growing tillers. When these side tillers were treated, the reverse movement took place. Similar observations for reciprocal movement have also been recorded in young plants of Lolium multiflorum at the two- and three-tiller stages of development by Sagar and Marshall (1966) and Marshall (1967). Forde (1966a) also reported this phenomenon in Cynodon dactylon (L.) Pers. It appears therefore that reciprocal movement between tillers does take place in many grasses and that the stage of development is an important factor which governs this type of movement. This movement seems to be restricted mainly to the young stages of growth when there is a considerable interdependence of the various parts of a tillering grass plant.
During the later stages of growth, from the six-tiller to the mature vegetative stages, intertiller translocation decreased markedly. Although the main tiller continued to supply small amounts of assimilates to growing regions of young tillers closely associated with it, the movement of assimilates from the daughter tillers ceased. Translocation from the daughter tillers was restricted mainly to their own growing regions and to young tillers closely associated with and on the same side as the treated tiller. It appears therefore that in E. curvula tillers become increasingly independent from the six-tiller stage onwards.
Tiller independence in other perennial grasses has been attained at about the seven-tiller stage in Lolium perenne (Forde, 1966b) and at the four-tiller stage in L. multiflorum (Sagar and Marshall, 1966; Marshall, 1967). It is interest- ing to note that there was extensive intertiller translocation in E. curvula at the four-tiller stage. Williams (1964) noted that in Phleum pratense, movement of assimilates from one tiller to another did not occur if the tillers were large. Similar observations were made in rice by Shen (1960). Quinlan and Sagar (1962) found that tillers of spring wheat became independent after ear emergence, while in winter wheats, Lupton (1966) found that they became autotrophic earlier.
Williams (1964) suggested that tiller independence may be related to the stage of differentiation of the vascular strands. This led Forde (1966a) to conclude that tiller independence in C. dactylon was probably attained only when tillers formed their own nodal roots.
However, evidence for the free movement of ions between tillers throughout the plant, tends to contradict the apparent individuality of the tiller. Troughton (1960) and Marshall (1967) showed that *?P was translocated to the whole of the plant after absorption by a single root of L. perenne and L. multiflorum
26 Journal of South African Botany
respectively. Williams (1960) confirmed this for °S in P. pratense.
In addition, defoliation experiments have added to the controversy. Mar- shall and Sagar (1965) showed that partial defoliation of tillers in L. multiflorum could induce export of assimilates from a normally independent tiller to damaged tillers. These authors therefore suggested that the plant must be regarded physiologically as a whole, in spite of the apparent individuality of the tiller.
The conclusion that can be drawn from these findings is that tillers become independent, once they have developed sufficient leaf area for the supply of all the assimilates needed for their own growth, and have established efficient rooting systems for the absorption of water and mineral salts. Under conditions of stress, probably caused by defoliation or shortage of nutrients, a reintegration of independent tillers is possible.
In the present investigation, faint images of mature leaves appeared on many autoradiographs. Similar observations were made by Hale and Weaver (1962) in the grapevine. Quinlan and Sagar (1962) reported that in wheat, many mature leaves became labelled when a young leaf assimilated CO,, and Hoshino, Nishimura and Okubo (1964) reported that there was some trans- location into old leaves of Ladino clover when a newly opened leaf was treated.
These findings are in contrast to work carried out by other investigators. Thaine, Ovenden and Turner (1959) demonstrated that in soyabean, import into mature leaves was exceptional. Williams (1964) and Sanderson and Sivapalan (1966) observed no significant movement into mature leaves of P. pratense and tea plants respectively.
The possibility that the labelling of mature leaves in wheat occurred via the transpiration stream following migration of label from phloem to xylem, as in the case of translocated herbicides, was investigated by Quinlan and Sagar (1962). However, no indication of xylem transference was obtained. In the grapevine, Hale and Weaver (1962) noted that the activity in mature leaves was not present in the veins but was restricted to the intervenal region of the laminae. This observation suggested that radiocarbon probably reached the mature leaves in the gaseous form. Subsequent investigations by these authors, confirmed this. It is possible therefore, that a similar phenomenon was respons- ible for the labelling of mature leaves in E. curvula and probably also in wheat and in Ladino clover. Microautoradiographic evidence to be presented in another paper demonstrates that CO, may be distributed within the plant in the gaseous form.
ACKNOWLEDGEMENTS The authors are grateful to the Council for Scientific and Industrial Research for financial assistance and to Mr. S. Kasavan for help with photography.
Translocation of Assimilates in Eragrostis Curvula Di
REFERENCES
Crafts, A. S. AND YAMAGUCHI, S., 1964. The autoradiography of plant materials. California Experimental Station Manual 35.
Forpbe, B. J., 19664. Translocation in grasses. 1. Bermuda grass. N.Z. Jnl. Bot. 4: 479-495.
, 19668. Translocation in grasses. 2. Perennial ryegrass and couch grass. N.Z. Jnl. Bot. 4: 496-514.
HAs, C. R. AND WEAVER, R. J., 1962. The effect of developmental stage on direction of translocation of photosynthate in Vitis vinifera. Hilgardia 33: 89-131.
HosHtIno, M., NiIsHIMURA, S. AND OxkuBo, T., 1964. Studies on the assimilation and trans- location of ##C in Ladino clover. Proc. Crop Sci. Soc. Japan 33: 130-134.
LittLe, E. C. S., 1962. Improvements in autoradiographic techniques for the study of the movements of carbon-14 labelled compounds in plants. Med. biol. Illust. 12: 31-34.
Lupton, F. G., 1966. Translocation of photosynthetic assimilates in wheat. Ann. appl. Biol. 57: 355-364.
MARSHALL, C. AND SAGAR, G. R., 1965. The influence of defoliation on the distribution of assimilates in Lolium multiflorum Lam. Ann. Bot. N.S. 29: 365-370.
MARSHALL, C., 1967. The use of radioisotopes to investigate organization in plants, with special reference to the grass plant. In: Isotopes in plant nutrition and physiology, 203-216, Vienna: I.A.E.A.
QUINLAN, J. D. AND SAGAR, G. R., 1962. An autoradiographic study of the movement of 14C-]abelled assimilates in the developing wheat plant. Weed Res. 2: 264-273.
SAGAR, G. R. AND MARSHALL, C., 1966. The grass plant as an integrated unit—some studies of assimilate distribution in Lolium multiflorum Lam. Proceedings of the Ninth International Grassland Congress 493-497.
SANDER,ON, G. W. AND SIVAPALAN, K., 1966. Translocation of photosynthetically assimilated carbon in tea plants. Tea Q. 37: 140-153.
SHEN, G. M., 1960. Translocation and distribution of assimilates from the leaves of the rice plant during its various developing periods. Acta Agric. Sin. 11: 30-40. [From Biol. Abstr. 35(19): 54003 (1969) ].
THAINE, R., OVENDEN, S. L. AND TURNER, J. S., 1959. Translocation of labelled assimilates in the soybean. Aust. J. biol. Sci. 8: 354-368.
TrouGHTON, A., 1960. Growth correlations between the roots and shoots of grass plants. Proceedings of the Eighth International Grassland Congress 280-283.
WILLIAMS, R. D., 1960. Nutrient uptake by grass roots. Proceedings of the Eighth International Grassland Congress 283-286.
WILLIAMS, R. D., 1964. Assimilation and translocation in perennial grasses. Ann. Bot. N.S. 28: 419-426.
YAMAGUCHI, S., AND Crafts, A. S., 1958. Autoradiographic method for studying absorption and translocation of herbicides using C-labelled compounds. Hilgardia 28: 161-191.
JIS. Afr. Bot. 41 (1): 29-36 (1975)
A QUALITATIVE STUDY OF THE NODULATING ABILITY OF LEGUME SPECIES: LIST 3
N. GROBBELAAR AND BRENDA CLARKE
(Margaretha Mes Institute for Plant Physiology and Biochemistry, University of Pretoria)
ABSTRACT
184 species of legumes indigenous to Southern Africa were examined for root nodules. Eleven of these (two belonging to the Mimosoideae, seven to the Caesalpinioideae and two to the Papilionatae) were consistently found to be without root nodules. Of the listed species, 141 had apparently not previously been examined for their ability to produce root nodules.
UITTREKSEL
’"N KWALITATIEWE STUDIE VAN DIE VERMOE VAN PEULPLANTSOORTE OM WORTEL-KNOPPIES TE VORM: LYS 3
184 peulplantsoorte wat inheems aan suidelike Afrika is, is vir die aanwesigheid van wortelknoppies ondersoek. Op elf spesies (2 lede van die Mimosoideae, sewe van die Caesal- pinioideae en twee van die Papilionatae) is deurgaans geen knoppies aangetref nie. 141 van die gelyste plantsoorte is skynbaar nog nie van te vore vir wortelknoppies ondersoek nie.
INTRODUCTION
Apart from reports by Mostert (1955), Grobbelaar et al. (1967) and Grobbe- laar and Clarke (1972), little is known about the nodulating ability of the 1 400-odd legume species (Phillips, 1951) that are indigenous to the Republic of South Africa, South West Africa, Botswana, Lesotho and Swaziland.
In an attempt to remedy this state of affairs a long-term survey was initiated at this laboratory some years ago and the present paper presents information obtained since the previous report was published by Grobbelaar and Clarke (1972).
PROCEDURE
The procedure was identical with that published earlier in greater detail (Grobbelaar et al. 1967). Where possible, plants were examined in the field for root nodules. In other cases plants were grown from seed in pots containing nitrogen-poor media and their roots were examined periodically for nodules. If nodules did not form within a reasonable time, the substrate was enriched with Rhizobium suspensions from available cultures and/or with soil from the natural habitat of the species when this was possible.
Accepted for publication 23rd April, 1974. 29
30 Journal of South African Botany
Annuals which failed to nodulate in at least three consecutive pot trials and perennials which did not nodulate within at least two years, were regarded as not being able to nodulate under local conditions.
In the case of species that were grown from seed in the laboratory, it was generally noted whether the germination was epigeal or hypogeal and whether a plumular hook developed during germination in the dark. Herbarium voucher specimens were prepared of all species tested. The specimens were identified at the Botanical Research Institute, Pretoria and deposited in the herbarium of the Department of General Botany, University of Pretoria.
Prof. and Mrs. Allen of the University of Wisconsin, U.S.A., have been compiling a card index of all available data regarding the nodulation of legumes for many years. The results of the present investigation are therefore regularly transmitted to them and they in turn indicate through personal correspondence whether other reports (published or personal) are available regarding the nodulating ability of the species concerned.
RESULTS AND DISCUSSION
The species on which information is provided are listed in Table 1. The present list contains only species which are indigenous to the Republic of South Africa, South West Africa, Botswana, Lesotho and/or Swaziland.
Of the 184 species listed in Table 1, eleven were consistently found to be without root nodules. Two of these (Acacia ataxacantha and Entada spicata) belong to the Mimosoideae and apparently were not previously examined for nodulation. Seven of the eleven species from the Caesalpinioideae that were examined consistently lacked root nodules. Of these Burkea africana, Bauhinia kirkii and Cassia petersiana apparently were not previously examined for nodulation. At least one negative report on the nodulation of the other four species was recorded by the Allens previous to the present study. These reports pertained to Schotia brachypetala, Bauhinia tomentosa, Cassia tomentosa and Cordyla africana. Only two of the species listed under the Papilionatae were consistently found to be without root nodules. These were Calpurnia aurea ssp. aurea and Virgilia divaricata, neither of which appears to have been examined previously for nodulation. It is instructive to note that of the 184 species listed in Table 1, 141 species apparently were not previously examined for nodulation. Of the 43 species about which the Allens did have earlier reports, it is only our report on Baphia massaiensis ssp. obovata which does not agree with the earlier reports. In this particular case only one earlier report (unpublished) reached the Allens and this report suggested that Baphia massaiensis ssp. obovata probably does not normally nodulate.
Of the 46 species for which germination characteristics are provided in Table 1, 43 were found to be epigeal. Of these 26 produced a plumular hook
Study of nodulating Ability of Legume Species: 3 31
whilst 8 did not. Two of the 3 species that germinated in a hypogeal fashion formed a plumular hook.
TABLE 1.
List of Indigenous Legume species examined for Nodulation.
Herbarium Nodulation Germin- Plant species? specimen ation? number? Present Allen’s study* records? MIMOSOIDEAE
Ingeae Benth. Albizia adianthifolia Cee ) en pc 25287 aig (C A. brevifolia Schinz. . eae 16941 +c C E+ AN, NOES eN, 56 o o 5» o o 6 16507 +¢ C E+ Acacieae Benth. Acacia ataxacantha DC. . ... . 16968 —c,f C E+ A@iborleaeiBurttDavysn ee ene 16914 +C (C E— A.erubescens Welw. . . ... . 17737 +c C E— A.exuvialis Verdoorn . . ... . 19436 =+-C C E+ @tieckaitSChinZiemeennnet nen 17420 +c A E+ AecerrandivBenthy ae en 16510 +c (C E A. grandicornuta Gesner . . . . . 17728 + C C E+ A. kraussiana Meisn. ex Benth. . 19169 +f Cc A. luederitzii Engl. var. retinens (Sim. )
Ross & Brenan. . ... . 16509 = C E AsmellenVerdoorms 5 5 5) 4 2. 17436 +c C E+ JN, WEIS Ole 5 6 6 o oo 5 Oo 16216 +-C A E— A. permixta Burtt Davy. . . . . . 17735 ape C E— A. rehmanniana Schinz . eS ior ge 17734 +e C E— A. senegal (L.) Willd var. leiorhachis
Brenan. . 17725 + C E+ A. senegal (L.) Willd var. “rostrata Brenan . 17733 a C E+ A. swazica Burtt Davy . . . . . . 17738 aC C E— A. tenuispina Verdoorn Be ares wh oe 17746 ae C E+ A. xanthophloea Benth.. . . . . . 17730 +¢ A E— Adenanthereae Benth. Dichrostachys cinerea (L.) Wight & Arn.
ssp. africana Brenan & Brummit var.
africana sens. lat. . 19168 +c C E+ D. cinerea (L.) Wight & Arn. ssp. nyassana
(Taub.) Brenan . . 23755 +e C Piptadenieae Benth. Entada spicata (E. Mey.) Druce . . . 16885 —c,f Cc E+
CAESALPINIOIDEAE
Dimorphandreae Benth. Burkea africana Hook.. . . . . . 16232 —c C E— Ambherstieae Benth. Schotia brachypetalaSond.. . . . . 16492 —c B E+ Bauhinieae Benth. Bauhiniaykirkii@livee 5 4 2 17552 —c Cc 1, WOmeEMOA IL, 6 6 6 0 60 0 o 86 25652 —c B E Cassieae Benth. Cassia _ biensis Breve). Mendonca &
WORE 5 oo ae 22819 +f A C. comosa Vogel. UE Gra se whit sie 20725 Spit A
32 Journal of South African Botany
Herbarium Nodulation Germin- Plant species specimen ation® number? Present Allen’s study? records? Cassia kirkii Oliv. 19397 +e (© E C. petersiana Bolle .. 16924 —c C C. plumosa (E. ee Voeel 20732 +f C C. tomentosa L. 5 25660 c B Swartzieae DC Cordyla africana Lour.. . 23851 —c B H PAPILIONATAE
Sophoreae Spreng Calpurnia aurea (Ait.) Benth. ssp. aurea . 20632 —c C E— Virgilia divaricata R. S. Adamson 19365 —c C Baphia massaiensis Taub. ssp. obovata
(Schinz.) Brummitt na) stat cle 25841 +e B Podalyrieae Benth. Cyclopia falcata (Harv.) Kies . 24381 +f C C. genistoides (L.) R. Br. 22953 +f Cc C. maculata (Andr.) Kies 16597 +f, ¢ C E C. montana Hofm. & Phill. 22907 +f C Podalyria cuneifolia Vent 17743 =-C C P. myrtillifolia Willd 16813 =f C Genisteae Bronn. Pleiospora cajanifolia Harv. 19191 +f C Lotononis azurea Benth. . . 23876 +f C L. calycina Benth. var. hirsutissima Duem-
mer. A ae ne cee 25304 +f (C L. cytisoides Benth. 22964 +c C E+ L. depressa Eckl. & Zeyh. 23653 fi C L. divaricata Benth. 24830 +f C L. macrosepala Conr. 25290 +f C L. mucronata Conr. 24831 +f C L. orthorrhiza Conr. . : 24834 +f C L. platycarpa (Viv.) Pichi—Serm. . 23622 +f C L. sericoflora Dimm. . : 25307 +f C L. stipulosa (Bak. f.) Schreiber 23623 +f A L. tenella Eckl. & Zeyh. 23731 +f C Pearsonia propinqua Duemmer. . 16233 +¢ C E+ P. sessilifolia (Harv.) Duemmer ssp. “fili-
folia (Bolus) Polhill . ; 25303 +f C P. sessilifolia (Harv.) Duemmer ssp. mar-
ginata (Schinz.) Polhill . 27337 +f C Aspalathus angustifolia Cay ) R. Dahler.
ssp. angustifolia . 23872 +f Cc A. cinerascens E. Mey... 3 23608 +f C A. laricifolia Berg. ssp. laricifolia . 16984 +f C A. triquetra Thunb. 27225 +f C Buchenroedera lotononoides Scott- Elliot . 27336 +¢ C E+ Dichilus pilosus Conr. 17745 +c C Melolobium adenodes Eckl. & Zeyh. . 28901 +f C M. obcordatum Harv. . . 28915 spilt C Crotalaria allenii Verdoorn 27334 ahi C C. brachycarpa Burtt Pas 23846 +f C C. burkeana Benth. 5 25253 sit C C. doidgeae Verdoorn 27863 +f C C. dura Wood & Evans. 27839 +f C C. erimicola Bak. f. . 19428 +c C
Study of Nodulating Ability of Legume Species: 3 33
Herbarium Nodulation Plant species! specimen ation® number? Present Allen’s study® records?
Crotalaria hispida Schinz . 25282 C C. laburnifolia L. ssp. australis (Bak, t)
Polhill. . a ip 19182 A C. maxillaris Klotzsch . . 25656 A E+ C. monteiroi Taub. ex Bak. f. var. galpinii
UME IDENAY 56 o oo «o o 6 20639 C E+ @anatalitiayMeisnas se oe ee eee 23890 A C. ochroleuca G. Don. Se an ae nee 19495 A E @hpisicarpaswelwaes.) =o a ee 19380 Cc CapodocarpayD Caan cn 23660 A G@erectarsteud sa ieee nner 16802 A C, adam, 5 5 «6 6 6 o 6 o 19381 C C. spartea Baker es ee ener ee an 20621 C E C. vasculosa Wallich . . ... . 23871 A C. virgultalis Burch. 3 22839 C Argyrolobium collinum Eckl. & t Zeyh. : 23717 Cc A. humile Phillips . . 3 23865 C A. lancifolium Burtt Davy . ee ae 22804 Cc A. megarrhizum H. Bol. . . . . . 27851 C A. pilosum Harv. os 28906 C A. stipulaceum Eckl. & Zeyh. gna” alo 19172 A A. transvaalense Schinz : 16795 C Loddigesia oxalidifolia (Sims) Phillips : 23679 C E+ Trifolieae Bronn. Trifolium africanum Ser. var. africanum . 18281 A Loteae Benth. otusidiscolor EaMeya a) ea ee 16811 C IL, ING IDESE 6 56 o 5 o 6 17649 A Galegeae Bronn. Indigofera acutisepala Conrath. . . . 23843 Cc I. adenocarpa E. Mey. . 16963 C I. costata Guill. & Perr. ssp. gonioides
(Hochst. ex Bak.) Gillett . . 20600 C I. costata Guill. & Perr. ssp. macra CE.
Mey.) Gillett. . 5 19187 C I. cuneifolia Eckl. & Zeyh. roa! ae te 23723 C I. cylindrica DC. . Se coor we 21676 c I, Giilocnina le IMSS 5 oo 6 o o « 28357 C I. foliosa E. Mey. . Beane We aye 24383 C I. glaucescens Eckl. & Zeyh. A eit us ee 27341 C I. hispida Eckl. & Zeyh. . . . . . 27342 C I. ingrata N. E. Br. eae Pa te 22885 C aplongebarbatayenel yee ee teee 23855 C [eemischocarpa Schiym a oe 23709 C I. mollicoma N. E. Br. Ae Ga eS oct 23733 Cc Ik, PRU INC IE IR 56 5 6 o 0 o 23875 Cc I. pongolana N. E. Br. 51 Ac eae 23644 C Ik, THAIN IER, 56 «o 6 6 6 6 o 25300 C Ik, Goaitevat ING I Be. 1 0 «© 0 6 « 23612 C Il; GEES IDS, 5 56 56 5 5 23720 C I, Gonai@h yea 6 o o 4 o o o 27223 C I. spinescensE. Mey. . .... . 23708 Cc [ey strictay of. 3 23678 Cc I. subulata Vahl. ex Poir. var. subulata . 28362 c
34 Journal of South African Botany
Plant species
Indigofera tinctoria L.
I. torulosa E. Mey. .
I. velutina E. Mey.. .
I. vicioides Jaub. & Spach. var. ‘vicioides . ;
Tephrosia aequilata Bak. ssp. australis
Brummitt ..
. capensis (Jacq. ) Pers. var. hirsuta Harv.
. contorta N. E. Br. ff
. cordata Hutch. & Burtt Davy ‘
. elongata E. Mey. var. elongata ee
. elongata E. Mey. var. pubescens Harv. .
. longipes Meisn. var. longipes
. lupinifolia (Burch.) DC. :
pseudocapitata H. M. Forbes ;
. Purpurea Pers. ssp. pvtostachya (DC.)
Brummitt : on es
. rhodesica Bak. f. .
. Villosa (L.) Pers.
. vogelii Hook. f. .
Sesbania cinerascens Welw. ex Bak.
S. coerulescens Harms . ;
S. sphaerosperma Welw.
S. transvaalensis Gillett.
Lessertia affinis Burtt Davy.
L. stenoloba E. Mey.
L. stricta Bolus .
Hedysareae DC.
Smithia erubescens (E. Mey.) Bak. f.
Desmodium natalitium Sond. . .
D. salicifolium DC. pane
Alysicarpus rugosus (Willd) De ssp. perennirufus J. Leonard : Neate
Alysicarpus zeyheri Harv.
Dalbergieae Bronn.
Dalbergia melanoxylon Guill. & Perr.
D. obovata E. Mey. .
Lonchocarpus capassa Rolfe
Phaseoleae Bronn.
Dumasia villosa DC...
Glycine wightii (Wight & Arn.) Verdc. ssp. wightii var. longicauda (Schweinf. ) Verdc. : re
Neorautanenia amboensis Schinz...
N. ficifolia (Benth. ex Hary.) C. A. Sm.
Erythrina caffra Thunb..
E. lysistemon Hutch.
E. zeyheri Harv.. . :
Canavalia maritima (Aubl. ) ‘Thouars j
Fagelia bituminosa (L.) DC. .
Rhynchosia ciliata (Thunb.) Druce :
R. cooperi (Harv. ex Bak. f.) Burtt Dany
R. longiflora Schinz... ;
R. minima (L.) DC. var. minima . 5
R. pentheri Schltr. var. hutchinsoniana Burtt Davy + “Siete aoe
Han 45355555345
Herbarium Nodulation Germin- specimen ation® number? Present Allen’s
study? records*
21680 +¢ A E+ 23632 Tale (C
25650 api C
19435 +c A
16933 SEI Cc
28927 +f €
27224 +f C
27227 +f Cc
20908 +f Cc
28383 +f C
23756 +c A E+ 23842 +f Cc
27833 +f C
19450 +e A Boe 23663 Sofi A
19430 +e A E 16929 {eC A
19185 +¢ A E+ 19189 +c A E+ 19389 +e Cc
16564 +c (Cc
23700 +f Cc
18240 +c C
23857 +c Cc
19190 +f Cc
25659 = fi A
19478 +c¢ A
25265 iat C
24374 +f A
19163 Sit A
20737 +f Cc
16500 +e C
16803 +f C
19431 +c A E 19371 +f C
23841 +f C
23764 +e A
22478 +e A
25842 +¢ C
20728 af A
17553 +f ©
22871 +f Cc
25274 =f C
23648 Sit C
28925 +f A
22903 +f C
Study of Nodulating Ability of Legume Species: 3 35
Herbarium Nodulation
Plant species! specimen ation? number? Present | Allen’s study® records?
Rhynchosia sublobata (Schum.) Meikle 22505 +¢ Cc H+ Eriosema cordatum E. Mey. var. cordatum 19180 +f C Flemingia grahamiana Wight & Arn. . 16938 +f Cc Vigna angustifoliolata Verdc. 21480 +f Cc V. davyi Bol. 25658 +f C V. decipiens Harv. . 25299 +f € V. luteola (Jacq.) Benth. 28361 +f A V. oblongifolia A. Rich. 19429 +c C E+ V. retusa Walp... . 27836 +f A V. unguiculata (L. ) Walp. ssp. dekindtiana
(Harms.) Verde... 3 25649 +f (S Dipogon lignosus (L.) Verde. . 19481 +c A H+ Dolichos angustifolius Eckl. & Zeyh. « 23707 +f Cc D. falciformis E. Mey. : ; 25844 +f (C D. hastaeformis E. Mey. . 23694 +f C D. trilobus L. subsp. transvaalensis Verde. 16583 fi A Macrotyloma maranguense (Taub. Verdc.) 21492 _ Cc
1. Species are arranged alphabetically within genera. The genera are arranged according to the system of de Dalla Torre & Harms (1963).
2. Voucher specimens are kept in the herbarium of the Dept. of General Botany, University of Pretoria.
3. A‘‘+” indicates that nodulation was observed and a “‘—”’ that nodulation was not observed during an observation period of at least two years. Plants studied at the laboratory are indicated by a “‘c’’ whereas species examined in the field are indicated by an “f”’.
4. The letters A, B and C refer to information obtained by private communication from Professor O. N. Allen, Department of Bacteriology, University of Wisconsin, U.S.A. A—nodulation previously observed B—species investigated previously but nodulation never observed C—species apparently not investigated previously for nodulation
5. “E” denotes epigeal and ‘‘H’’ hypogeal germination. The ‘‘+”’ or “‘—”’ indicates whether a plumular hook is formed during germination or not.
ACKNOWLEDGEMENTS
The financial assistance provided by the Department of Agricultural Tech- nical Services and the South African Council of the International Biological Programme is gratefully acknowledged.
REFERENCES
ALLEN, E. K. AND ALLEN, O. N., 1961. The scope of nodulation in the leguminosae. In: Recent Advances in Botany, 585-588. Univ. of Toronto Press.
De DALLA Torre, C. G. AND HARMS, H., 1963. Genera siphonogamarum ad systema engleria- num conscripta. Wiesbaden: Wissenschaftliche Neudrucke.
De Souza, D. I. A., 1966. Nodulation of indigenous Trinidad legumes. Trop. Agric., Trin. 43: 265-267.
GROBBELAAR, N. AND CLARKE, B., 1972. A qualitative study of the nodulating ability of legume species: List 2. J] S. Afr. Bot. 38: 241-247.
GROBBELAAR, N., VAN BEYMA, M. C. AND Topp, C. M., 1967. A qualitative study of the nodulating ability of legume species: List 1. Publs. Univ. Pretor ia, New Series 38.
Mostert, J. W. C., 1955. Observations on the nodulation of some leguminous species. Fig. Ss Afr., 30: 338-340.
Puitures, E. P., 1951. The genera of South African flowering plants. Second Edition. Mem. bot. Surv. S. Afr. 25.
JIS. Afr. Bot. 41 (1): 37-38 (1975)
ANDREAEA NITIDA HOOK. f. & WILS., A NEW RECORD OF A SOUTHERN HEMISPHERE MOSS FOR AFRICA
B.A: €. L. E. SCHELPE (Bolus Herbarium, University of Cape Town)
ABSTRACT
Andreaea nitida Hook. f. & Wils., previously known from South America, Tristan da Cunha S. E. Australia, Tasmania, New Zealand, Auckland Islands and New Guinea is recorded from mountain habitats in the south western Cape Province.
UITTREKSEL
ANDREAEA NITIDA HOOK. f. & WILS., °*N NUWE VINDPLEK VIR ’N MOS IN DIE SUIDELIKE HALFROND VAN AFRIKA
Andreaea nitida Hook. f. & Wils., voorheen bekend van Suid-Amerika, Tristan da Cunha, S. O. Australié, Tasmanié, Nieu-Seeland, Auckland Eilande en Nieu-Guinee, is ook in berg habitatte in die Suidwes-Kaapprovinsie gevind.
Sim (1926) in his “Bryophyta of South Africa’? recorded only two species of the dark coloured montane moss genus Andreaea from South Africa. Of these two, the one “‘costate” species, i.e. with a median longitudinal nerve in the leaf, was A. subulata Harv., the type locality of which is Table Mountain on the Cape Peninsula. More recently, the present author (Schelpe 1969) pointed out the existence in South Africa of another “‘costate’’ species of Andreaea similar to the temperate South American A. subenervis Hook. f. & Wils. and to the New Zealand 4A. nitida Hook. f. & Wils. On the strength of Sainsbury’s (1955) description and comments, the present author had ascribed some of the South African material to A. nitida, but on the publication of Schultze-Motel’s (1970) monograph of the costate species of Andreaea, it became clear that the name A. nitida (syn. A. subenervis) was applicable to specimens from the mountains of the south western Cape Province enumerated as follows:
CAPE PROVINCE—Cape Peninsula: Table Mountain, Silverstream Ravine, on flat rock wet from dripping water, 3 000 ft., 11.xii.1955, E. Ester- huysen 25110 (BOL). Ceres: Hex R. Mountains, Roodeberg, rock surface with water trickling over, S. E. aspect, 6 500 ft., 27.xii.1952, E. Esterhuysen 20964 (BOL). Clanwilliam: Koue Bokkeveld Mountains, Hexberg, on rock in dry rocky watercourse below summit plateau, 5 000 ft., 24.11.1951, E. Esterhuysen 18475 (BOL). Ladismith: Swartberg, Toverkop, rock face, seasonal water- course on cliffs, E. side, 6 500 ft., 16.xii.1956, E. Esterhuysen 26823 (BOL).
Accepted for publication 9th December, 1974. 37
38 Journal of South African Botany
Tulbagh: Great Winterhoek, rock face on small falls in seasonal watercourse dry in summer, S. aspect, 5 000 ft., 27.1.1957, E. Esterhuysen 27044 (BOL).
Consequently, these South African records fill a hitherto apparent gap in the Southern Hemisphere distribution of A. nitida given by Schultze-Motel (1970) ranging through South America from Colombia through Ecuador to Bolivia, to Tristan da Cunha, south eastern Australia, Tasmania, New Zealand, Auckland Islands and north eastern New Guinea.
Among the costate species of Andreaea in South Africa, A. nitida is dis- tinguished by its ovate leaves with a rather broad indistinct nerve, from A. subulata which has subulate, nerved leaves from a broader base. Although only recorded by Sim (1926) from Table Mountain, A. subulata is now known to occur on a number of the south western Cape mountain ranges (Cedarberg, Great Winterhoek, Matroosberg, Villiershoek, Wemmershoek, Du Toits Kloof, Hottentots Holland and Swartberg), as well as having a wide Southern Hemi- sphere distribution as indicated by Schultze-Motel (1970) from South America (Colombia to Tierra del Fuego), Falkland Islands, Africa (S. W. Cape moun- tains and Ruwenzori), Madagascar, Borneo (Mt. Kinabalu), S. E. Australia, Tasmania, New Zealand and its subantarctic islands.
REFERENCES SAINSBURY, G. O. K., 1955. A Handbook of the New Zealand Mosses. Bull. R. Soc. N.Z. 5: 1-490.
SCHELPE, E. A. C. L. E., 1969. Three new records of Southern Hemisphere Bryophyta for South Africa. (Africa) J/ §. Afr. Bot. 35: 109-112.
SCHULTZE-MOTEL, W., 1970. Monographie der Laubmoosgattung Andreaea I. Die costaten Arten. Willdenowia 6: 24-110.
Sim, T. R., 1926. The Bryophyta of South Africa. Trans. R. Soc. S. Afr. 25: 1-475.
JIS. Afr. Bot. 41 (1): 39-40 (1975)
A SHORT NOTE ON COMBRETUM EDWARDSITI EXELL *R. L. VERHOEVEN AND **H. P. VAN DER SCHIJFF (Department of General Botany, University of Pretoria)
ABSTRACT
Combretum edwardsii is a liane with interxylary phloem in the secondary xylem. C. edwardsii was described by Exell (1968) from sterile material. The type specimen was collected by D Edwards. Fertile material collected by H. P. van der Schijff is described.
UITTREKSEL
KORT AANTEKENINGE OOR COMBRETUM EDWARDSI EXELL
Combretum edwardsii is ’n liaan met interxilére floéem in die sekondére xileem. C. edwardsii is deur Exell (1968) van steriele materiaal beskryf. Die tipe monster is deur D. Edwards versamel. Fertiele materiaal versamel deur H. P. van der Schijff word beskryf.
Combretum edwardsii is a woody climber with stalked elliptical opposite leaves. Brown scales are conspicuous on the leaves and both the petioles and young stems are covered with unicellular combretum hairs.
The leaves are bifacial with anomocytic stomata in the abaxial epidermis only. The scales are about 90-100 »m in diameter and the head consists of 9 to 14 marginal cells. The total number of cells of the head varies from 14 to 25 (Fig. 1A). In a longitudinal section the scale consists of a stalk of four cells and a flat multicellular uniseriate head (Fig. 1B).
The anatomy of the stem and root is the same as the general structure of the family Combretaceae with interxylary phloem in the secondary xylem. In the
epidermis
50 um
A
Fic. 1. Scale of Combretum edwardsii. A. Surface view; B. Longitudinal section.
* Department of Botany, University of the Orange Free State, Bloemfontein. ** Department of General Botany, University of Pretoria. Accepted for publication 4th April, 1974.
39
40 Journal of South African Botany
other three climbers, viz. C. bracteosum (Hochst.) Brandis, C. microphyllum Klotzsch, and C. mossambicense (KI.) Engl., interxylary phloem is absent. Of all the South African species Combretum edwardsii is therefore the only climber with interxylary phloem. The latter three species are also the only South African representatives with stalked glands on the leaves (Verhoeven & Van der Schijff, 1973).
C. edwardsii Exell was described by Exell (1968) from sterile material in the hope that fertile material would be collected at a later stage. The type specimen was collected by D. Edwards (Edwards 3147, PRE, holotype). Un- fortunately the collecting date was not given.
Material of C. edwardsii was collected by R. L. Verhoeven on 6th December, 1968 in Natal, District Impendle about 64 kilometres from Pietermaritzburg on the road to Bulwer. From this material the anatomy of the plant was studied. The species is a definite climber and occurs abundantly in the area.
The first fertile material and so far the only known, was collected by H. P. van der Schijff at Mariepskop in the Eastern Transvaal on the 25th October, 1962. The specimen, viz. Van der Schijff 6273, was wrongly identified as Quis- qualis parviflora Gerr. and was only recently rediscovered and correctly identi- fied as Combretum edwardsii Exell.
At Mariepskop it grows as a robust liane with a basal stem diameter of up to 20 cm thick, in dense indigenous mountain forest. It tops the tree stratum and the cream-coloured inflorescences are found in open sunlight on the forest canopy—probably the main reason why fertile material was not collected at an earlier stage.
Inflorescence a spike 3 to 4 cm long; Sepals 4 (3,5 mm); Petals 4 (1,5 mm); Stamens 8 (7-8 mm); Pistil 8,5 to 9,5 mm; Style 5,5 to 6 mm. The unilocular inferior ovary bears at its top a short floral tube. At the top of the floral tube are four sepals. Alternating with the sepals are four petals. The eight stamens are inserted on the floral tube. Interior to the bases of the filaments the floral tube is developed into a lobed hairy disk.
REFERENCES EXELL, A. W., 1968. Notes on the Combretaceae of Southern Africa. Bolm Soc. broteriana, Ser. 2, 42: 5-33.
VERHOEVEN, R. L. AND VAN DER SCHIIFF, H. P., 1973. A key to the South African Combretaceae based on antomical characteristics of leaf. Phytomorphology 23: 65-74.
JIS. Afr. Bot. 41 (1): 41-45 (1975)
BATCHELOROMYCES, A NEW GENUS OF ANNELLIDIC DEMATIA- CEOUS HYPHOMYCETES ON PROTEA CYNAROIDES IN SOUTH AFRICA
W. F. O. MARASAS
(Plant Protection Research Institute, Pretoria)
P. S. VAN WYK AND P. S. KNOx-DAVIES
(Department of Plant Pathology, University of Stellenbosch)
ABSTRACT
A description of Batcheloromyces Marasas, Van Wyk & Knox-Davies, a new genus of dematiaceous Hyphomycetes is given. The new genus is characterized by mycelial plugs in the stomata and superficial mycelium forming pulvinate radiating plates. The conidiogenous cells (annellides) are calyciform with distinct annellations and one- or two-celled, brown, bacilliform conidia are produced by successive percurrent proliferations of the conidiogenous cells. A single species, B. proteae sp. nov. on Protea cynaroides L. is included in the genus at present.
The relationships of Batcheloromyces with other genera of dematiaceous Hyphomycetes are discussed.
UITTREKSEL
BATCHELOROM YCES, ’N NUWE GENUS VAN ANNELLIDIESE DONKERKLEUR- IGE HIFOMISETE OP PROTEA CYNAROIDES IN SUID-AFRIKA
Batcheloromyces Marasas, Van Wyk & Knox-Davies, ’n nuwe genus van donkerkleurige Hifomisete word beskryf. Die nuwe genus word gekenmerk deur miseliumproppe in die huidmondjies en oppervlakkige miselium wat kussingvormige miseliumplate vorm. Die koni- diogene selle (annellides) is koppievormig met duidelike annellasies en een- of twee-sellige, bruin, bacillusvormige konidiums word deur opeenvolgende proliferasies van die konidiogene sel deur die opening wat gelaat word deur die afskeiding van die vorige konidium, gevorm. Slegs ’n enkele spesie B. proteae sp. nov. op Protea cynaroides L. word tans in die genus geplaas. Die verwantskappe van Batcheloromyces met ander genera van Hifomisete word bespreek.
In the course of an investigation of foliicolous fungi of South African Proteaceae (Van Wyk, 1973), an interesting dematiaceous Hyphomycete was collected on Protea cynaroides in the Cape Province. This fungus differs from all the fungi that have been described on Proteaceae and from the known genera of dematiaceous Hyphomycetes (Ellis, 1971). Consequently it is described here as Batcheloromyces proteae gen. et sp. nov. The type specimen is deposited in the Mycological Herbarium of the Plant Protection Research Institute, Pretoria, Republic of South Africa (PREM).
Batcheloromyces Marasas, Van Wyk et Knox-Davies gen. nov.
Deuteromycotina, hyphomycetes. Coloniae orbiculatae, radiantes, atrae, maculas rufobrunneas efficientes. Mycelium partim in stomatibus immersum, partim superficiale, ex hyphis ramosis, septatis, brunneis, ab stromatibus
Accepted for publication 22nd January, 1974. 41
42 Journal of South African Botany
Fics. 1-7. Batcheloromyces proteae, type material (PREM 44850). 1, Colony on leaf of Protea cynaroides showing the black, pulvinate plates of radiating mycelium, X4; 2, Calyciform conidiogenous cells on superficial mycelium. Uncoated material on leaf examined under a Coates and Welter Model 100-4 field emission scanning electron microscope, X8500; 3, Section through leaf of P. cynaroides showing the mycelial plug in the stoma and the superficial plate of radiating mycelium on the leaf surface above the stoma, X900; 4, Annellide with three annellations and an undetached conidium, X2500; 5, Annellides with two successively formed conidia clinging together in false basipetal chains, X2500; 6, Calyciform conidiogenous cells (annellides) produced as lateral erect or ascending branches on the superficial hyphae, X2000; 7, One-celled, brown, bacilliform conida, X1300.
Batcheloromyces, New Genus of Hyphomycetes 43
pulvinatis ramosis compositum. Conidiophora macronemata, mononemata, simplicia, brunnea, solitaria ex lateribus hypharum oriunda. Cellae conidiogenae monoblasticae, in conidiophoris incorporatae, terminales, brunneae, calyci- formea, percurrentes. Conidia singulatim in apice conidiophori oriunda, solitaria sed interdum in catenis falsis curtis persistentia, sicca, brunnea, crassitunicata, laevia, ellipsoidea vel bacillaria, continua vel uniseptata.
Species typica: Batcheloromyces proteae Marasas, Van Wyk et Knox-Davies
Colonies on leaves circular, radiating, black in colour, causing reddish- brown leaf spots. Mycelium in the leaf tissue restricted to the stomata as stro- matic mycelial plugs. On the leaf surface the mycelium forms pulvinate, sporo- dochium-like plates of radiating olivaceous-brown hyphae above the stomata and a network of delicate, branched, septate, superficial hyphae closely adhering to the leaf surface. Conidiophores macronematous, mononematous, simple, brown, arising as short, erect or ascending branches of the superficial hyphae, usually composed of a single terminal conidiogenous cell. Conidiogenous cells (annellides) monoblastic, integrated, terminal, brown, calyciform, percurrent with distinct annellations. Conidia arise solitarily as blown-out ends of the apex of the conidiogenous cell, dry, brown, smooth, thick-walled, ellipsoidal or bacilliform with rounded ends or a truncate base and rounded apex, one- celled or with a single transverse septum; conidia produced by successive percurrent proliferations of the conidiogenous cell may hang together in short fragile, false basipetal chains.
Etymology: The genus is named after Mr. F. C. Batchelor of Stellenbosch on whose farm the type species was collected.
Batcheloromyces proteae Marasas, Van Wyk et Knox-Davies sp. nov.
Coloniae amphigenae, usque ad 2 cm diam. Stromata superficialia, pulvinata, 110-250 4m diam. Hyphae 2,2-4,2 »m crassae. Condiophora effusa sed creber- rima in centro coloniae, plerumque ex cella conidiogena una composita. Cellae conidiogenae calyciformes, usque ad 3 proliferationes successivas (= annel- lationes) producentes, 3,0-5,3 4m longae, 3,3-4,8 »m latae ad apicem, 2,2-3,1 ~m latae ad basim. Conidia solitaria vel usque ad 3 conidia interdum in catenis falsis fragilibus persistentia, brunnea, laevia, crassitunicata, ellipsoidea vel bacillaria, continua vel uniseptata, 3,9-9,1 x 2,8-4,2 um.
Habitat in foliis vivis Proteae cynaroides L., Stellenbosch, Cape Province, Republic of South Africa, 15 viii. 1973, Van Wyk, PREM 44850, holotypus.
Colonies on leaves amphigenous, circular, radiating, up to 2 cm diam, discrete but becoming confluent and covering large areas of leaf surface, black, causing a reddish-brown discolouration of the leaf tissue (Fig. 1) which may be visible in the corresponding areas of the opposite leaf surface. Mycelium
44 Journal of South African Botany
in the leaf tissue composed of thick, dark brown, stromatic mycelial plugs in the stomata (Fig. 3), but no subcuticular or intra-epidermal mycelium could be seen. The mycelial plugs are erumpent through the stomata and form black, pulvinate, sporodochium-like plates of radiating hyphae above the stomata (Fig. 3) and a network of delicate superficial hyphae closely adhering to the leaf surface. Each colony comprises numerous black superficial mycelial plates, 110-250 um diam (Fig. 1) and these plates are composed of a single layer of radiating, septate, branched, brown hyphae, 2,2-4,2 um diam. Conidiophores arise solitarily as short, simple, erect or ascending, lateral branches on the superficial hyphae (Fig. 2,6), effuse but concentrated in the centre of the colonies directly above the stomata, usually consisting of a single terminal conidiogenous cell (Fig. 2,6) but sometimes the conidiogenous cell proliferates through the conidial scar to produce a second conidiogenous cell or a hypha which can presumably give rise to a second conidiogenous cell at a higher level. Conidio- genous cells (annellides) are brown, integrated, terminal, percurrent with up to three annellations (Fig. 4), typically calyciform, 3,0-5,3 wm long, 2,2-3,1 um diam at the base and widening towards the apex which is 3,3-4,8 »m wide (Fig. 2,6), sometimes doliiform, 3,0-4,6 wm high and 2,3-3,8 ~m wide. Conidia arising singly as blown-out ends of the apex of the conidiogenous cell (Fig. 4,6), solitary, or conidia produced by successive percurrent proliferations of the conidiogenous cell may hang together in fragile, false basipetal chains of two or three conidia (Fig. 5), dry, brown, oblong-ellipsoidal or bacilliform with both ends rounded or truncate at the base, smooth, thick-walled, one-celled (Fig. 7) or rarely with a single transverse septum and then not constricted at the septum, 3,9-9,1 <2,8-4,2 wm.
On living leaves of Protea cynaroides L., Protea Heights, Devon Valley, Stellenbosch district, Cape Province, 15 August 1973, Van Wyk (PREM 44850, holotype).
Specimens examined: CAPE—Protea Heights, Stellenbosch, 15 August 1973, Van Wyk (PREM 44850, holotype); Assegaaibos, Jonkershoek, Stellen- bosch, 23 August 1972, Van Wyk (PREM 44852); Oudebos, Betty’s Bay, 15 March 1973, Van Wyk (PREM 44851); George, 25 September 1973, Van Wyk (PREM 44916).
The annellations on the conidiogenous cells are so conspicuous under the light microscope (Fig. 4) that we can only assume that this fungus belongs in Section III of Hughes’ original classification (Hughes, 1953b) and that conidial development is annellidic as defined by the “First International Specialists’ Workshop-Conference on Criteria and Terminology in the Classification of Funge Imperfecti” (Kendric, 1971). If this interpretation is in fact correct, Batcheloromyces appears to be related to the form-genera Deightoniella Hughes, Pollaccia Baldacci & Ciferri, Spilocea Fries and Oedothea Sydow as defined
Batcheloromyces, New Genus of Hyphomycetes 45
by Hughes (1953a) and Ellis (1971). The vegetative mycelium and calyciform annellides are also similar to those of the genus Phragmospathula as described by Subramanian and Nair (1966). The combination of morphological charac- teristics of the fungus on P. cynaroides, i.e. mycelial plugs in the stomata, superficial plates of radiating mycelium, calyciform annellides and brown, thick-walled, bacilliform, one- or two-celled conidia, clearly differetiates it from all these apparently related genera. In the absence of electron microscopic studies of conidium ontogeny, it is not possible to draw final conclusions as to the relationships of Batcheloromyces at present.
ACKNOWLEDGEMENTS
The authors are indebted to Dr. M. B. Ellis of the Commonwealth Mycolo- gical Institute, Kew for his examination of a portion of the type collection and for helpful comments.
REFERENCES
ELuis, M. B., 1971. Dematiaceous Hyphomycetes. Kew: Commonwealth Mycological Institute.
Huaues, S. J., 1953a. Some foliicolous Hyphomycetes. Can. J. Bot. 31: 560-577.
Huaues, S. J., 1953b. Conidiophores, conidia and classification. Can. J. Bot. 31: 577-659.
KENDRICK, B., (ed.), 1971. Taxonomy of Fungi Imperfecti. Toronto: Univ. of Toronto Press.
SUBRAMANIAN, C. V. and Nair, N. G., 1966. Panchanania and Phragmospathula, two new genera of the Hyphomycetes. Antonie van Leeuwenhoek—J. Microbiol. Serol. 32: 381-386.
VAN WYK, P. S., 1973. Funguspatogene van die genera Protea, Leucadendron en Leucospermum met spesiale verwysing na Phytophthora cinnamomi, Ph. D. Thesis. Stellenbosch: Univ. of Stellenbosch
JIS. Afr. Bot. 41 (1): 47-50 (1975)
PLANTAE NOVAE AFRICANAE “Ex Africa semper aliquid novi’’—Pliny. SERIES XXXIII
R. H. COMPTON (Compton Herbarium, Kirstenbosch)
ABSTRACT
Six new species of flowering plants from Swaziland are described.
UITTREKSEL Ses nuwe blomplantsoorte vanaf Swaziland word beskryf.
INTRODUCTION
The following names and descriptions are given for new species collected in Swaziland and continue the publication in this periodical, vol. 33, pp. 293-304, October 1967. The specimens cited are preserved in the Herbarium of the Botanical Research Institute, Pretoria, and the Compton Herbarium, National Botanic Gardens, Kirstenbosch. My sincere thanks are due to the heads and staffs of these institutions for their invaluable assistance throughout my work on the flora of Swaziland.
Kalanchoe alticola Compton (Crassulaceae).—Herba perennis erecta succulenta glabra, leviter glauca. Folia pauca opposita, in caulis parte inferiore sessilia, ovata, integra, basim contracta, obtusa, marginibus subcoriaceis. Inflorescentia in pedunculo elongato terminalis, dichasio-corymbosa; calycis segmenti acuti; corolla flava, base ventricosa, segmentis patentibus acutis.
Perennial, glabrous throughout. Rhizome horizontal, shallow, elongated, c 5 mm diam., prolonged at apex into erect aerial shoot c 25 cm high, occa- sionally with lateral shoots also. Shoot bears up to 5 pairs of imbricating somewhat succulent slightly glaucous leaves on its basal half, above which it bears one or two remote pairs of reduced leaves separated by long internodes and ends in a dichasio-corymbose inflorescence c 5 cm diam. Lower leaves sessile, narrowing gradually to base, widest above middle, obtuse, up to 4 cm long x2 cm wide with entire subcoriaceous margin. Leaves on inflorescence axis similar, up to 1,5 cm long x1 cm wide. Calyx c 5 mm long, lobes acute. Corolla tube inflated, c 1 cm long, lobes spreading, c 6 mm long, pointed, clear light or deep yellow.
Accepted for publication 21st October, 1974. 47
48 Journal of South African Botany
Hab. Swaziland. Highveld, peaty rock surfaces and crevices. Fl. May— September.
2631 (Mbabane): Mukusini Hills c 1 300 m alt. (-AB), Compton 32107 (NBG, holotype); Compton 30437, Karsten s.n., 11/5/1963.
Lotononis spicata Compton (Leguminosae).—Caules erecti, dense foliosi, ubique sparse pilosi, pauci. Folia sessilia, trifoliata, segmentis subequalibus erecto-patentis, ovato-lanceolatis, obtusis. Inflorescentia dense spicata, floribus in foliorum superiorum axillis, foliis obtectis. Calycis lobus anterior angustus, superiores minuti, reflexi. Corolla flava; vexillum obtusum, alae angustiores, carina obtusa.
Rhizomatous perennial. Shoots erect, up to 25 cm high, with few branches near base. Stems, leaves and calyces with scattered long fine hairs. Leaves numerous, hiding stem, trifoliate, sessile, erecto-patent, lobes about equal, lanceolate or ovate-lanceolate, entire, obtuse, up to 2 cm long <7 mm wide, usually less. Flowers borne along upper half of shoot forming a dense spike, shorter than and partly hidden by subtending leaves. Calyx c 5 mm long to the tip of the very narrow anterior lobe, other calyx-lobes very small, reflexed. Corolla yellow, vexillum a little darker, erect, obtuse, c 8 mm long; alae narrow, c 4 mm long; carina obtuse, c 8 mm long.
Hab. Swaziland. Highveld grassland. Fl. March. 2631 (Mbabane): Mpala- leni, c 1 300 m alt. (-AB), Compton 32111 (NBG, holotype); 3 miles W. of Mankaiana (-CA), c 1 300 m alt. Compton 28666; Ngwenya Mts. c | 650 m alt., Compton 26737.
Acalypha moggii Compton. (Euphorbiaceae).—Radix lignosa, perennis. Caules flexuosi, internodiis elongatis, foliis paucis, ubique sparse pilosi. Petiola perlonga; lamina late ovata, acuta, forte cordata, marginibus dentatis. Planta monoecia. Inflorescentiae axillares, inferiores masculae, longe pedunculatae, floribus dense longe spicatis. Inflorescentia feminea lata, haud pedunculata; bracteae latae, laciniatae, imbricatae, floribus obtectis, stigmatis haud productis.
Woody rootstock from which rise single shoots up to 130 cm high, un- branched except near apex. Stem prominently ribbed, sparsely hairy, up to 7 mm diam., internodes up to 10 cm long. Leaves alternate, petiole up to 8 cm long, lamina broadly ovate, up to 8 cm long x6 cm wide, tapering to narrow point, deeply cordate, margins regularly closely dentate, teeth up to 3 mm long and wide, veins very narrow. Monoecious. Inflorescences axillary, males below, females usually single near apex of shoot, occasionally lower. Male peduncle slender, up to 5 cm long, sometimes with leafy basal bract, flowers very numerous, densely crowded in slender spike up to 7 cm long x4 mm diam., perianth 2 mm long, hairy, stamens protruding. Female spike clasped at base by small leaves, massive, up to 8 cm long x2,5 cm diam., usually less;
Plantae Novae Africanae: Series XX XIII 49
bracts broad, imbricate, laciniate, acuminate, strongly ribbed, with copious stalked glands; stigmas not evident; capsule c 4 mm diam.
Hab. Highveld swamps. Fl. January-April. 2430 (Acornhoek): Rosehaugh, Pilgrim’s Rest, 1 200 m alt. (-DD), Mogg 13666 (PRE, holotype); Belfast, 2000 m alt., Bruce 500. 2531 (Komatipoort): Piggs Peak, 1 000 m alt. (-CC), Compton 27625, 30615.
First collected by Dr. A. O. D. Mogg, 10th January 1938.
Dicliptera divaricata Compton (Acanthaceae).—Caules subrigidi herbacei, ramis numerosis divaricatis, internodiis elongatis. Folia opposita, sparse puberula, breviter petiolata, lanceolata, minute apiculata. Flores paucae in axillis superioribus; corolla pallide purpurea.
Erect, rather stiffly herbaceous, up to c 60 cm high, branching freely at c 45°. Stems slender, prominently 4-ridged, almost or quite glabrous, internodes of main stem up to 10 cm long. Leaves in opposite pairs, spreading or deflexed, very sparsely puberulous throughout; petiole up to 5 mm long; lamina lanceo- late, broadest below middle, minutely apiculate, faintly veined, up to 4 cm long x 1 cm wide. Flowers rather few in upper leaf-axils, opening singly; outer bracts c 8 mm long x2 mm wide, long-attenuate; calyx 4 mm long; bracts and calyx minutely puberulous; corolla mauve, tube c 8 mm long, lobes c 1 cm long; filaments inserted at mouth of corolla, straight, c 8 mm long, anther- lobes roundish borne at different levels.
Hab. Swaziland. Lowveld. Fl. March. 2731 (Louwsberg): Ingwavuma Poort, 100 m alt. (-BB), Compton 28614 (NBG, holotype); Gollel, 100 m alt. (-BD), Rubin s.n.
Lobelia coddii Compton (Campanulaceae).—Planta perennis, fastigiata, caulibus rectis. Folia multa, sessilia, erecto-patentia, mesonevro prominente, marginibus pauco-dentatis. Flores multae in foliarum reductarum axillis, pedicellatae; calyx campanulatus, lobis expansis, angustis, apiculatis; corolla azurea vel variegata, segmentis dorsalibus reflexis, ventralibus expansis.
Tufted perennial, producing a large number of erect straight shoots from a woody rootstock, up to c 45 cm high, sparingly erectly branched near base, more widely branched near apex. Stems and leaves thinly hairy, becoming glabrous with age. Stems with few narrow ribs. Leaves numerous, rather uniform in shape and size throughout, erecto-patent, sessile, lanceolate, with strong midrib, minutely apiculate, margins entire or with one or two forward- directed pointed teeth on each side, up to 8 mm long x4 mm wide, seldom larger. Flowers numerous, mostly solitary in axils of reduced leaves; pedicel up to 2 cm long; calyx 7 mm long, hairy, campanulate at base, lobes spreading, narrow apiculate; corolla blue, sometimes with yellow blotch, two upper lobes 2 mm long, reflexed on either side of united anthers, three lower lobes expanded, obtuse, c 3 mm long x2 mm wide.
50 Journal of South African Botany
Hab. Swaziland. Rocky places on mountains. Fl. about March. 2631 (Mbabane): Hills ¢ 3 miles W of Mbabane, rocks, c 1 100 m alt. (-AC), Compton 25727 (NBG, holotype); Hills c 3 miles W of Mbabane, shallow soil on granite crest c | 200 m alt., Codd 9516 (PRE).
Lobelia delicatula Compton (Campanulaceae).—Planta delicatissima. Caules tenues, undulati. Folia alternata, breviter petiolata; lamina late triangulata, obtusa, fragilis, marginibus subcrenulatis; calycis segmenti angusti, acuminati; corolla alba vel pallide azurea.
Very slender annual, sometimes perennating with a short rootstock. Roots few. Stems up to c 15 cm long, undulate, c 1 mm diam., ridged, with few slender erect hairs. Leaves alternate, irregularly spaced on stem; petiole up to 5 mm tong; lamina very delicate, equilateral-triangular, up to 1 cm long and wide, base almost straight, sides curved, margins slightly crenulate, apex obtuse, lower surface with a few minute hairs. Upper leaves smaller and narrower with few axillary flowers. Pedicel capillary, up to 2 cm long, flower horizontal becoming erect on withering. Calyx 4 mm long, segments narrow, acuminate. Corolla pale blue or white, c 6 mm long, upper lip c 2 mm shorter than lower which has 3 short lobes c 1 mm wide.
Hab. Swaziland. Highveld, moist peaty soil in shade of rocks or on moist banks in forest shade. Fl. February to April. 2631 (Mbabane): Hill N.E. of Mbabane, c 1 400 m alt. (-AC), Compton 27538 (NBG, holotype); Compton 25809; Black Mbuluzi Falls, c 1 100 m alt. (-AA), Compton 26852.
Book REVIEWS
FLORA OF LESOTHO, by A. Jacot Guillarmod, with 474 pages and map of Lesotho on inside cover. Lehre, Germany: J. Cramer. DM150 (R30,85).
Before the publication of the work presently under review, the only treatment of Lesotho plants was E. P. Phillips’ “A Contribution to the Flora of the Leribe Plateau and Environs”’, published in the Annals of the South African Museum in 1917. However, his 338 page tax- onomic treatment, incorporating descriptions of some new species, included not only the Lesotho collection of Dieterlen and others, but also those from the bordering mountainous areas in Natal, the Orange Free State and the eastern Cape Province.
Since then, Dr. Amy Jacot Guillarmod has collected extensively in Lesotho and has searched European herbaria for specimens collected in that country. The results of her praiseworthy activity both in the field and in herbaria are embodied in the major part of the publication under review. This constitutes a valuable source of information, especially in regard to the location of specimens, but the frequent reference of specimens to an unidentified specific category underlines how much critical taxonomic work on the Lesotho flora still remains to be done.
In addition to interesting sections on Lesotho history, descriptive ecology, land tenure and usage, biographical notes on collectors of Lesotho plants, indices to Lesotho vernacular names and local uses, there is a valuable index of Lesotho place names giving map references. However, despite the wealth of information to be gleaned from this work, the title given to this book, at least to a taxonomist, is rather a misnomer.
E. A. SCHELPE.
TREES OF THE KRUGER NATIONAL PARK, VOL. 2, by P. van Wyk, with pp vit-285—597, 104 full page colour plates. Cape Town: Purnell and Sons, 1974. R17,50.
The second and final volume of Mr. P. van Wyk’s Trees of the Kruger National Park has now appeared, covering the families Euphorbiaceae to Compositae in the Engler system. It follows the same format and style as the previous volume and one cannot but express satis- faction that the high standards of colour reproduction that characterised the plates in volume I, continue to be maintained. To photograph a mature tree in the lowveld environment, showing its characteristic adult growth form, without being obscured by the associated vege- tation, and to present it as a glamorous illustration, requires a skilful balancing of priorities, not to mention the arduous footslogging involved in seeking out photogenic specimens. In the case of novelties like Steganotaenia araliacea, Hymenodictyum parvifolium and numerous others, it is more than likely that Pieter van Wyk’s photographs are the first colour illustrations of these species ever published.
Clearly, the author has made a great effort to keep one step ahead of that apparently invincible monster—Nomenclatural Change. He has succeeded insofar as it is ever possible to be up to date with one’s nomenclature and provides the reader with the most recent name changes as well as the latest taxonomic decisions for the Kruger Park tree flora. New records for the Kruger Park, too recent for inclusion in the format of the main text, are also listed. In addition to an index of scientific names with full author citations, there is an index of common names in English, Afrikaans and eight Bantu languages. Regrettably, the chapter headings of these two indexes have been transposed. Several other careless slips in major headings (Glu- siaceae for Clusiaceae) are irksome but not serious. They could surely have been corrected by the publishers.
Many readers will, I am sure, be gratified to learn that the author has adopted a con- servative approach as to what constitutes a tree. He has omitted a number of shrubby plants whose growth forms or stature have been either inflated or etiolated by some authorities in order to accord them tree status in the latest edition of the National List of Trees. His sane approach is to be greatly commended. (cf. the note under Myrsinaceae).
It must be stressed that these two volumes are not field guides, to be toted along when motoring through the Kruger Park. Notwithstanding their durable bindings, they are books for the library to be consulted and enjoyed at leisure. Each volume is priced at R17,50. The reader can be assured that by present day standards, this represents a bargain, not easily
equalled. LPR . P. ROURKE.
31
JOURNAL OF SOUTH AFRICAN BOTANY SUPPLEMENTARY VOLUMES
Volume Price
1(1944) | THE GENUS OXAL/S IN SOUTH AFRICA, by
T. M. Salter, with pp. 355, 10 plates, 73 figs.
(out of print) 2(1954) THE GENUS MURALTIA, by Margaret R.
Levyns, with pp. 247, 2 plates, 37 maps, 88
figs. (out of print) 3(1959) THE GENUS BABI/ANA, by G. Joyce Lewis,
with pp. 149,33 plates(somecol.) . . R3,00 4(1965) THE GENUS AGAPANTHUS L'HERITIER, by
Frances M. Leighton, with pp. 50, 14 plates
(some col.), 3 maps, 21 figs. . R3,00 5(1965) STUDIES OF THE SOUTH APRICAN SPECIES
OF EULOPHIA, by A. V. Hall, with pp. 248,
1 plate, 43 maps, 41 figs. . R3,00 6(1966) AN ANNOTATED CHECK UST OF THE
FLORA OF SWAZILAND, by R. H. Compton,
with pp.191,2 maps . . R3,00 7(1969) TAXONOMIC STUDIES ON SOROCEPHA-
LUS R.BR. AND SPATALLA SALISB., by
J. P. Rourke, with pp. 124, 10 plates,
33 maps, 21 figs. . R4,00 8(1972) TAXONOMIC STUDIES ‘ON LEUCOSPER-
MUM R.BR., by J. P. Rourke, with pp. 194,
ilatigsae R6,00 9(1972) THE GENUS ROMULEA IN SOUTH AFRICA,
by Miriam P. de Vos, with pp. 307, 23 maps,
97 figs. (some col.) . R8,00 10(1972) GLADIOLUS: A REVISION OF THE SOUTH
AFRICAN SPECIES, by G. Joyce Lewis,
A. Amelia Obermeyer with T. T. Barnard,
with pp. 316, 32 col. plates: 52 esis: Pale
Town, Purnell . . ; R8,50
Note:
To avoid loss through exchange rates etc., it would be appreciated if Overseas payments to this Institution could be made by BANK DRAFT in South African currency only.
Orders for Supplementary Vol. No. 10 should be addressed to: Purnell & Sons. (S.A.) (Pty.) Ltd. P.O. Box 4501. Cape Town, South Africa.
Supl. Vols. No. 3—9 available from: National Botanic Gardens of S.A. Kirstenbosch, Newlands, 7700. South Africa.
JIS. Afr. Bot. 41 (2): 53-62 (1975)
NOTES ON PROTEA IN SOUTH AFRICA: THE PROTEA TENA X TANGLE
J. P. ROURKE (Compton Herbarium, Kirstenbosch)
ABSTRACT
The correct usage of the name Protea tenax (Salisb.) R. Br. is indicated and the variability of this sues discussed. The taxon previously referred to as‘P. tenax (Salisb.) R.Br. var. latifolia Meisn/, is accorded specific rank and redescribed as P. foliosa Rourke.
UITTREKSEL
AANTEKENINGE OOR PROTEA IN SUID-AFRIKA: DIE PROTEA TENAX VER- WARRING
Die korrekte gebruik van die naam Protea tenax (Salisb.) R.Br. word aangedui en die veranderlikheid van die soort bespreek. Die takson voorheen bekend as P. tenax (Salisb.) R. Br. var. latifolia Meisn., word verhef tot ’n soort en herbeskryf as P. foliosa Rourke.
INTRODUCTION
Protea tenax (Salisb.) R.Br. as construed by Phillips and Stapf (1912), contained two elements: a decumbent narrow leaved form and an erect, shrubby broad leaved form, the latter being referred to as Protea tenax (Salisb.) R.Br. var. latifolia Meisn. Due to the citation of only one specimen of the decumbent narrow leaved form and the citation of numerous collections of the erect, broad leaved form in their account of Protea tenax in the Flora Capensis, the name P. tenax came to be associated with the erect, broad leaved form in most herbaria. The position was not improved by the erroneous citation of a Niven collection of the decumbent narrow leaved form of P. tenax under Protea canaliculata Andr. resulting in the association of the name P. canaliculata with material that should have been referred to P. tenax.
A careful study of the type material of all the names involved as well as several years of field work, have now clarified the position. This note is an attempt to clear up the confusion of the past and to define the taxa in question, preparatory to publishing a revision of Protea in South Africa.
DISCUSSION
Field studies revealed that two series of allopatric populations are involved which can easily be distinguished by differences in growth habit.
1. From van Stadensberg eastwards to Grahamstown, a relatively uniform taxon occurs, having a shrubby growth habit with erect, usually unbranched
Accepted for publication 21st November, 1974. 53
54 Journal of South African Botany
stems to 1,5 m in height, very prominent, lanceolate, oblanceolate or elliptic leaves and small globose inflorescences. In most herbaria, material of this taxon has been named P. tenax (Salisb.) R.Br. but a comparison with the type illustra- tion (Paradisus Londinensis tab. 70), clearly shows that it is net P. tenax. Although recognized as different and described by Meisner as P. tenax (Salisb.) R.Br. var. latifolia Meisn. this taxon must be accorded specific rank on the basis of its distinct morphology and geographical isolation. However, it is not possible to raise Meisner’s epithet to specific rank since this step would involve the creation of a homonym. It is therefore described de novo as Protea foliosa.
Protea foliosa Rourke, sp. nov.
A species cognatae differt, propter habitu erecto fruticoso, caulibus erectis non ramosis, foliis prominentibus glabris, lanceolatis vel late ellipticis, inflorescentias cingentes.
Frutex parvus, rotundatus, 1,5 m altus, caulibus numerosus rectus, caudicibus exorientes. Caules recti, non ramosi (tantum raro ramosi), glabri, 6-8 mm in diam. Folia lanceolata, oblanceolata vel late elliptica, 9-16 cm longa, 1,5—7 cm lata, obtusa vel acuta; glabra, sed pubescentia initio. /nflorescentiae ovoideae vel globosae, cyathiformes, 3-4 cm in diam., reconditae, foliis amplis superanti- bus. Receptaculum involucrale 1,5 cm in diam., late convexum. Bracteae involu- crales 4—5 seriata. Series intima oblonga spathulata, 3-3,5 cm longa, | cm lata, apices rotundatos concavos. Series externa ovata vel late ovata, obtusa, 0,7—1,5 cm longa, 0,7-1,2 cm lata; pagina externa sparsa sericea, glabrescens, viridi- flava, sordida-carmina suffusa. Perianthium 2,5-3 cm longum, glabrum, praeter perianthium limbum manifeste lanatum ad apicem. Perianthium limbi lineares, acuminati, glabri sed apices dense lanatos. Stylus 2,5 cm longus, arcuatus adaxialis, contractus terminalis. Stigma peranguste lineare, 5 mm longum, compressum, longistrorsum sulcatum. Squamulae hypogynae ovatae, acutae, 1,5 mm longae.
A rounded shrub to 1,5 m tall with numerous straight, erect stems arising from a woody subterranean rootstock. Stems stout, 6-8 mm in diam., erect, glabrous, rarely branched. Leaves lanceolate, oblanceolate or broadly elliptic, 9-16 cm long, 1,5—7 cm broad, narrowed in the petiolar region; apex acute to obtuse; margins occasionally undulate; coriaceous, surface minutely granulate, pubescent at first, soon glabrous. Inflorescences ovoid to globose, cyathiform, 3-4 cm wide hidden by large overtopping leaves; terminal but occasionally axillary. Involucral receptacle broadly convex, 1,5 cm in diam. Jnvolucral bracts 4-5 seriate; innermost series oblong-spathulate 3-3,5 cm long, 1 cm wide, apex broadly obtuse, concave; outer series ovate to broadly ovate, obtuse, 0,7-1,5 cm long, 0,7—1,2 cm wide; outer surface minutely fimbriate; usually greenish yellow, occasionally flushed with dull carmine, particularly along margins. Perianth
Notes on Protea in South Africa 55)
Fic. 1. Protea foliosa Rourke, part of the type material, Rourke 1410. Above: flowering shoot, x 0,5. Below: inflorescence at anthesis, life size.
56 Journal of South African Botany
straight to very slightly adaxially arcuate, 2,5-3 cm long; tube 8-10 mm long, glabrous; claws slender with strongly undulate margins, glabrous; limbs 8-10 mm long, linear, apices acuminate, glabrescent to glabrous for greater part of their length but apex thickly lanate with a white or brown indumentum. Style 2,5 cm long, straight to slightly adaxially curved. Pollen presenter linear, acute, 5 mm long. Ovary obovoid, 3-4 mm long covered with long tawny trichomes. Hypogynous scales ovate acute, 1,5 mm long.
Protea magnolifolia Buek in Drége, Zwei Pfl. Doc.: 118; 127; 213 (1843)—nom. nud.
Protea caulescens E. Mey. ex Meisn. in DC., Prodr. 14: 244 (1856)—nom. nud. Protea tenax (Salisb.) R.Br. var. /atifolia Meisn. in DC., Prodr. 14: 244 (1856); Phill. & Stapf in Fl. Cap. 5: 587 (1912). Type: Circa Galgebosch et van Stadens- berg, Drége s.n. in herb. Meisn. (NY, holo.!).
Two manuscript names, P. magnolifolia and P. caulescens have been applied to this species but both are nomina nuda. The varietal epithet, P. tenax (Salisb.) R.Br. var. Jatifolia Meisn. cannot be raised to specific rank as it would then become a homonym of P. /atifolia R.Br. Since there is no legitimate specific epithet available for this species, it is here described de novo.
Type Material: Van Stadensberg Forest Reserve, between Loerie and Otterford, 18.6.1974, J. P. Rourke 1410 (NBG holo.; PRE, STE, BOL, GRA, K, MO, S, M, C, iso.).
Diagnostic Characters: P. foliosa may be distinguished from related species on account of its shrubby, upright growth habit with erect, unbranched (or only very rarely branched) stems, and prominent glabrous leaves, lanceolate to broadly elliptic, which completely surround the small globose inflorescences.
Common in the Van Stadensberg mountains generally at elevations above 150 m, this species is also found on most Table Mountain Sandstone out- croppings around Uitenhage and Port Elizabeth. Other populations also occur in the fynbos on the hills around Riebeeck East and Grahamstown at elevations of up to 600 m, on soils derived from Witteberg Quartzite. The distribution range of P. foliosa falls within a region experiencing a mean annual rainfall of about 500— 750 mm, distributed more or less evenly throughout the year but with a slight winter maximum. Flowering takes place erratically between March and Septem- ber with the peak of the season being in May and June. However, odd in- florescences may be produced at almost any time of the year. A massive woody rootstock or lignotuber gives P. foliosa a high degree of fire resistance.
SPECIMENS EXAMINED
CAPE PROVINCE—33724 (Steytlerville): Assagaybosch (-CD), May, Zeyher 3667 (SAM, GRA).
Notes on Protea in South Africa 57
3325 (Port Elizabeth): Mountains N.W. of Uitenhage (-CB), Aug., Loots s.n. (NBG 19425); Van Stadensberg and Galgenbosch (-CC), 27.12.1829, Drége s.n. (P, SAM); Van Stadens, 6.5.1934, Long 1272 (PRE, GRA, RU, K); Van Stadensberg forest reserve between Loerie and Otterford, 18.6.1974, Rourke 1410 (NBG, PRE, STE, GRA, K, M, MO, S, BOL, C); Van Stadensberg, Oct.—Nov., Zeyher 388 (SAM, PRE); Van Stadensberg next to Galgebosch, 14.2.1814, Burchell 4755 (K); 25 km up the Elands River road (-CD), 3.8.1960, Acocks 21265 (NBG, PRE); Kamehs, Uitenhage, May, Paterson 142 (BOL, GRA); 10 km west of Port Elizabeth (-DC), May, Crump s.n. (NBG 82, 485); Bethelsdorp, near Port Elizabeth July, L. Bolus s.n. (NBG 82486); Bethelsdorp Kloof, Aug., Fries, Norlindh & Weimark 240 (S, NBG).
— 3326 (Grahamstown): Rocky mountains, Zwartbergen, Albany (-AA), Burchell 3436 (K); Between Riebeeck East and Grahamstown (-AB), July 1813, Burchell 3508 (K); Near Coldstream, Grahamstown (-BC), June, Pym 1174 (SAM, GRA); Prope Grahamstown, Mac Owan 1266 (BOL); Bottom of Signal
—
1 2 3 4 5 6 8
RiGy 2: Protea tenax (Salisb.) R.Br., showing variation in the leaf morphology. _ (1) Rourke 908; (2) Williams 1782; (3) Wurts 2081; (4) Rycroft 2336; (5) Oliver 4539; (6) Taylor 936; (7) Rycroft 2336; (8) L. Abel s.n. (NBG 62889). The scale line represents 1 cm.
58 Journal of South African Botany
Hill, Grahamstown, Aug., Daly & Sole 473 (GRA, K); Grassy mountains, Grahamstown, Galpin 26 (GRA); Featherstone Kloof, north side Grahamstown, 13.6.1963, Noel 433 (RU); Beggars Bush, P. Coetzee s.n. (GRA 1173); Mountain drive Grahamstown, Aug. 1926, L. L. Britten 5424 (PRE); Paynes farm, Bush- mans River Poort, Alexandria (-CB); 31.8.1954, Archibald 5624 (PRE, K).
2. The second series of populations ranges throughout the entire length ot the Longkloof on the north slopes of the Outeniqua mountains, from Wabooms- kraal in the west to Humansdorp in the east and also occurs in the Baviaans- kloof and Winterhoek mountains. All are characterised by having decumbent, trailing stems, often with secund or subsecund leaves and greenish-white, shallowly crateriform inflorescences, occasionally flushed with pink. On close examination, little variation in the dimensions or morphology of the in- florescences and floral parts is observable but marked differences in foliar characters give each population a distinctive facies. Considering the variety of habitats occupied, this variation is not surprising. Habitats vary from arid fynbos on hot north facing slopes where the mean annual rainfall is + 300 mm, to fairly dense sclerophyll in moist sheltered localities where the mean annual rainfall is -+ 770 mm.
Leaf width is the only significant variable within this series of populations. Populations occupying drier habitats tend to have narrow, linear, slightly glaucous leaves while populations with broader, elliptic or oblanceolate leaves, having little or no glaucescence, are found on moister sites. However, no satisfactory correlation between geographical distribution and leaf dimensions can be found. In general, the narrow leaved populations display very little range in leaf width but in the broader leaved populations the range in leaf width, even on a single flowering shoot, is considerable (Fig. 3). At first sight, the narrow leaved forms do appear to constitute a distinct group but their random occurrence among broader leaved populations as well as the continuous variation in leaf width (Figs. 2 & 3) makes the recognition of two taxa unwise, even at subspecific rank. Protea tenax (Salisb.) R.Br., the earliest legitimate epithet available for this species is typified by a fine colour plate depicting a specimen which most closely resembles a population growing in the vicinity of Kareedouw, eastern Longkloof.
Protea tenax (Salisb.) R.Br. in Trans. Linn. Soc. Lond. 10: 88 (1810); Meisn. in DC., Prodr. 14: 244 (1856); Phill. & Stapf in Fl. Cap. 5: 587 (1912). Erodendrum tenax Salisb., Parad. Lond. t. 70 (1807). Type: Plate 70 in Parad. Lond.
Protea longifolia Salisb., Parad. Lond. t. 37 (1806), non P. longifolia Andr. (1801). Type: Plate 37 in Parad. Lond.
Erodendrum longifolium (Salisb.) Salisb, ex Knight, Cult. Prot.: 46 (1809).
Notes on Protea in South Africa 59
Erodendrum foetidum Salisb. ex Knight in Knight, Cult. Prot.: 46 (1809)—nom. superfl. Type: As for Erodendrum tenax Salisb.
Protea canaliculata sensu R.Br. in Trans. Linn. Soc. Lond. 10: 88 (1810); sensu Meisn. in DC., Prodr. 14: 241 (1856); et sensu Phill. & Stapf in FI. Cap. 5: 591 (1912)—all according to specimens cited,—non P. canaliculata Andr.
Protea undulata Phill. in Kew Bull 1910: 233 (1910). Type: Without collector or locality (K, holo. !)
= 25 3 io Range in 45 20 10 ? Leaf width ME
Fic. 3. Continuous variation in the leaf width of Protea tenax (Salisb.) R.Br. (from 26 herbarium collections). Each horizontal stroke represents a single collection, showing the range in leaf width in that collection. This range is plotted by taking the widths of the broadest and narrowest leaves of each collection and linking these two points with a horizontal stroke.
60 Journal of South African Botany
Protea acaulos (L.) Reich. var. cockscombensis Archibald in Flower. PI. Afr. 28 t. 1115 (1951). Type: The Cockscomb, Winterhoek mts., Archibald 3435 (PRE, holo.!; NBG, GRA, iso.!)
A low decumbent shrub with trailing, sparsely branched stems arising from a woody rootstock. Stems to 2 m in length, usually trailing, occasionally suberect, 3-6 mm in diam., puberulous at first, soon quite glabrous, usually flushed with carmine. Leaves linear, narrowly lanceolate to lanceolate, elliptic or oblanceo- late, 8-18 cm long 2-28 mm wide, apices acute to obtuse; slightly twisted and decurrent at base; lamina flat or only occasionally with incurved margins. Inflorescences shallowly crateriform, opening widely, 4-6 cm in diam. Involucral receptacle broadly convex, 1,5-2 cm in diam. Involucral bracts 4—5 seriate; innermost series broadly oblong, obtuse, 3,5—4 cm long, 1—1,5 cm wide, the apices concave and rounded; outer series very broadly ovate 1-2 cm long, 1-2 cm wide, apices obtuse to rounded; outer surface sparsely sericeous becoming irregularly glabrous, margins minutely fimbriate; bracts greenish-yellow, sometimes flushed with carmine. Perianth straight to slightly adaxially arcuate, 2,5-3 cm long, glabrous except for the lanate apices of the perianth limbs; tube 8 mm long; claws hyaline, undulate; limbs 8 mm long, linear, apices acuminate, thickly lanate with a dense white indumentum. Sty/e 2,5-3 cm long, straight to slightly adaxially curved. Pollen presenter linear, acute, 5 mm long. Hypogynous scales irregularly ovate, 1 mm long.
Diagnostic Characters: The decumbent growth habit with trailing stems bearing leaves which are usually secund or subsecund (very rarely ascending) and wide, shallowly crateriform inflorescences, 4-6 cm across the mouth when open, distinguish P. tenax from related species.
Populations which are here assigned to P. tenax occur throughout the Long- kloof, on the Outeniqua, Baviaanskloof, Humansdorp and Groot Winterhoek mountains. Waboomskraal in the Outeniqua range and the Cockscomb near Uitenhage, are, respectively, the extreme westerly and easterly points of its distribution range. Ecologically, the habitats of P. tenax may range from arid fynbos where the mean annual rainfall is about 300 mm, to fairly dense fynbos where the mean annual rainfall is about 770 mm. At all localities it is confined to soils derived from Table Mountain Sandstone. Its altitudinal range extends from 250-1 750 m. Flowering is erratic but appears to be correlated with rainfall, which, although it is spread more or less evenly throughout the year within the distribution area of P. tenax, tends to have a slight winter maximum. Most of the inflorescences are produced between May and September although they can appear at almost any time of the year.
Notes on Protea in South Africa 61
under 500 metres
500 — 1000 »
tJ
=] 1000 — 1500
S500 — 2000 » en 1} iF ‘ K ae ; bo R foliosa over 2000 »
P tenax
Fic. 4. Distribution ranges of Protea tenax (Salisb.) R.Br. and P. foliosa Rourke.
SPECIMENS EXAMINED
CAPE PROVINCE—3322 (Oudtshoorn): Waboomskraal near top of Outeniqua Pass (-CD), 18.7.1954, Rycroft 1628 (NBG); Summit of old Duivels Kop Pass (-DC), 21.9.1967, Rourke 908 (NBG).
—— 3323 (Willowmore): Potjiesrivierhoogte, Uniondale (-CA), Rourke 363 (NBG); Potjiesrivierhoogte Pass, 14.8.1960, Acocks 21411 (PRE, NBG); 12 km S.W. of Uniondale on fynbos hills, 3.9.1951, Acocks 16026 (PRE); Rocky hill near Haarlem (-DB), 14.3.1814; Burchell 4977 (K); Walletjies, between Misgund and Joubertina, Longkloof (-DC), 10.7.1974, Bayliss 6417 (NBG); Louterwater, 3.5.1933, Compton 4211 (BOL); North slopes Tsitsikamma State Forest “The Heights’’, near Joubertina (-DD), 8.6.73, Kruger 1621 (STE). — 3324 (Steytlerville): Baviaanskloof Pass (-CB), 19.5.1971, Stayner s.n. (NBG 92541); Pisgoedvlakte, summit of Cambria Pass, Baviaanskloof, 13.4.1973, Williams 1782 (NBG); About 0,5 km north of Kareedouw village (-CD), 2.5.1960, Wurts 2081 (NBG); 3 km from National Rd., on Kareedouw Pass to Assegaai Bosch, 2.5.1960, Wurts 2078 (NBG); Kareedouw Pass, 28.10.1961, Rycroft 2336 (NBG); Summit of Pass north of Assegaai Bosch 25.3.1966, Williams 721 (NBG); Baviaanskloof Mts., above Couties Kraal, N.W. of Cambria (-DA), 12.9.1973, Oliver 4539 (STE); North slopes of Cockscomb (-DB), 7.10.1962, L. Abel s.n. (NBG 62889); Mac, Groot Winterhoek, 23.9.1953, H. C. Taylor 936 (NBG); The Cockscomb, Winterhoek mts., Archibald 3435 (PRE, GRA, NBG); The Cockscomb, 6.9.1954, Noel 434 (GRA); Rietvlei, Humansdorp (-DC), 11.11.1941, Esterhuysen 6623 (BOL).
62 Journal of South African Botany
— 3424 (Humansdorp): 28 km beyond Humansdorp on way to Clarkson (-AB), C. F. Hayes s.n. (GRA); Sandy soil on plains at Kromme River (-BA), Niven 34 (K).
REFERENCES
PHILLIPs, E. P. AND STAPF, O., 1912. Flora Capensis 5: 587 London: Lovell Reeve. RourkKgE, J. P., 1974. Notes on Protea in South Africa. JI S. Afr. Bot. 40: 291-300.
c.T.P. LTD.
JIS. Afr. Bot. 41 (2): 63-66 (1975)
IDENTIFICATION OF ALLANTOIN IN PROTEA COMPACTA SEED
S. E. DREWES AND J. VAN STADEN (Departments of Chemistry and Botany, University of Natal, Pietermaritzburg)
ABSTRACT
By means of I.R., N.M.R. and mass spectrometry it was shown that on a dry mass basis approximately two per cent of Protea compacta seed consists of allantoin.
UITTREKSEL
IDENTIFIKASIE VAN ALLANTOIN IN PROTEA COMPACTA SAAD Deur middel van I.R., K.M.R. en massaspektrometrie is vasgestel dat op ’n droé massa basis ongeveer twee persent van die saad van Protea compacta uit allontoien bestaan.
INTRODUCTION
While investigating the effect of endogenous hormones on the germination of Protea compacta seed (Brown and Van Staden, 1973) it was observed that concentrated ethanolic extracts of embryo material, when stored at 5 °C, yielded large quantities of needle-like crystals. On a dry weight basis the embryos contained approximately two per cent of this crystalline material. The crystals were insoluble in practically all organic solvents and only slightly so in water. They did, however, dissolve in hot water but crystallised rapidly upon cooling.
Due to the relative abundance of the crystalline substance and the fact that
it may play an important role in the physiology of the seed, its identification was undertaken.
MATERIAL AND METHODS
Embryo material of Protea compacta R.Br. was homogenized with 80% ethanol, using approximately 25 ml of ethanol per gramme dry mass of material, and was then shaken for 24 hours at 18 °C. After filtering the extract ina Buchner funnel, the residue was thoroughly washed with 80% ethanol. The combined ethanolic filtrates were concentrated to a small volume under vacuum at 35 °C. The concentrated extract was kept at 5° C for 48 hours during which period fine needle-like crystals formed. These crystals were recovered by filtration, purified by recrystallization from water and subsequently used for chemical identifica- tion. The crystalline material was optically inactive.
Accepted for publication 30th January, 1975. 63
64 Journal of South African Botany
SOLVENT
Fic. 1. N.M.R. spectrum of crystalline compound from Protea compacta in (CD3). SO.
Infra-red spectra were run as KBr discs, N.M.R. spectra were recorded on a Varian T60 instrument while a Varian CH7 instrument was used for mass spectrometry.
RESULTS
Using the above procedure every gramme of dry embryo material yielded 17 mg of a relatively pure crystalline compound which had a melting point of 235 °C. In water this substance showed no UV-absorption above 280 nm. Infra-red spectral analysis yielded the following significant data: vy» max 3438, 3340 (NH): 1780, 1705 (CO). The N.M.R. spectrum in (CDs), SO gives rise to:
(i) a singlet at 6 10.53 (single proton -CO-NH-CO-),
(11) a singlet at 5 8.05 (single proton -HC-NH-CO-),
(111) a doublet at 6 6.90 (single proton -CH-NH-CONH.), (iv) a singlet at 6 5.77 (two protons -CONA,), and
(v) a quartet at 6 5.26 (single proton -NH-CH-NH) (Fig. 1).
Low-resolution mass spectrometry of the unknown compound indicated a molecular ion of m/e 158. Major peaks were recorded at m/e 141 (M-17), 130 (M-CO), 115 (M-CONH), 87 (base peak), 60, 44 and 54 (Fig. 2).
65
RELATIVE ABUNDANCE 7,
ise [M*]
40 60 80 100 120 140 160
Fic. 2. Mass spectrum of crystalline compound from Protea compacta seed identified as allantoin.
Elemental analysis gave the following results:
Found Calculated C 30,48% 30,38 % gi BI 3,82% INGE 263) 35,44 %
The above analytical data yielded a molecular formula of C,H,N,O3. The physical data coupled with the fact that there was no depression of melting point when mixed with authentic material proves that the compound is racemic allantoin with the structure as indicated in Fig. 2.
DISCUSSION
The identification of allantoin in seed of Protea compacta adds to its already long history as a plant constituent. It was first isolated in 1881 by Schulze and Barbieri from sprouting shoots of Platanus orientalis and has subsequently been found in 22 families of flowering plants (Tracey, 1955). The fact that it is usually present in relatively high concentrations in plants suggests that it plays an important role in nitrogen metabolism. It would appear as if it is most abund- ant in xylem sap, ripening fruits, seeds and young developing leaves, flowers and seedlings (Tracey, 1955). All these are either organs in which large quantities of food reserves are stored or tissues which have a rapid rate of metabolism. Very
66 Journal of South African Botany
little definite information is, however, available regarding the physiological function of allantoin in plant material. Its frequent occurrence in seed led to the suggestion that it serves as a storage product for nitrogen (Mothes, 1958). It has also been postulated that it plays a role in the detoxification of ammonia, that it might be the form in which nitrogen is translocated in plants, or that it isa degradation product of purines (Mothes, 1958). The fact that it is present in Protea compacta seed where the major seed reserves are proteins (-- 65 °%), yet is absent in Leucospermum reflexum, a genus in the same family where the food reserve is lipid (+55°%) (unpublished), suggests that it may act as a storage product in the former species. Since allantoin levels have been reported to increase during germination it may be that it is actively synthesized or alterna- tively it is formed as a result of purine degradation during this process (Barash, 1972).
The possibility of allantoin acting as an intermediate storage produce in the seed of Protea compacta is presently being investigated.
ACKNOWLEDGEMENTS
The financial assistance of the Council for Scientific and Industrial Research, Pretoria, South Africa is gratefully acknowledged.
REFERENCES BarasH, J., 1972. Accumulation of urea and allantoin during purine utilization by germinating spores of Geotrichum candidum. J. gen. Microbiol. 72: 539-542.
Brown, N. A. C. AND VAN STADEN, J., 1973. The effect of stratification on the endogenous cytokinin levels of seed of Protea compacta and Leucadendron daphnoides. Physiol. PI. 28: 388-392.
Motes, K., 1958. Ammoniakentgiftung und Aminogruppenvorrat. In: W. Ruhland (ed.) Handbuch der Pflanzenphysiologie 8: 716-762. Berlin: Springer-Verlag.
ScHULZE, E. AND BARBIERI, J., 1881. Uber des Vorkommen von Allantoin im Pflanzenorganis- mus. Ber dt. chem. Ges. 14: 1602-1605.
Tracey, M. V., 1955. Urea and Ureids. In: K. Paech and M. V. Tracey (eds.) Modern methods of plant analysis 4: 119-141. Berlin: Springer-Verlag.
c.T.P.LTD.
JIS. Afr. Bot. 41 (2): 67-85 (1975)
STUDIES IN THE BULBOUS LILIACEAE IN SOUTH AFRICA: 5. SEED SURFACE CHARACTERS AND GENERIC GROUPINGS
J. P. JESSOP (Schonland Botanical Laboratory, Rhodes University)*
ABSTRACT
Scanning electron microscope studies were made on the surface of the seeds of 49 specimens of 25 genera of South African Liliaceae. Five groups of genera of bulbous Liliaceae are recognised on the basis of seed surface characters.
It is suggested that three tribes should be recognised in the South African bulbous Liliaceae—the Scilleae, Massonieae and Bowieae. These are defined and the genera in each are listed.
UITTREKSEL
STUDIES VAN DIE BOLDRAENDE LILIACEAE IN SUID-AFRIKA: 5. KENMERKE VAN DIE SAADOPPERVLAK EN GENERIESE GROEPERINGS Skandeer-elektronmikroskoop opnames van die saadoppervlak van 49 eksemplare van 25 genera Suid-Afrikaanse Liliaceae is gemaak. Vyf groepe genera van boldraende Liliaceae word erken op die basis van die saadoppervlakte kenmerke. Dit word voorgestel dat drie tribusse boldraende Liliaceae in Suid-Afrika erken behoort te word—Scillae, Massonieae en Bowieae. Hierdie word defineer en die genera in elk gemeld.
INTRODUCTION
The South African bulbous Liliaceae genera have been placed in various tribes of the Liliaceae by different authors. It is generally agreed that the majority belong to a single tribe. Krause (1930) called the major group Scilloideae. Schizobasis, Bowiea and Eriospermum were included in a separate tribe—the Eriosperminae.
Hutchinson (1959) recognised the Scilleae and two other tribes—the Mas- sonieae with two genera (Massonia and Daubenya) and the Bowieae with three genera (Schizobasis, Bowiea and Eriospermum).
Other authors have commented on the grouping of the bulbous genera, for example Phillips (1951) linked Massonia with Polyxena, Neobakeria, White- headia, Neopatersonia, Androsiphon and Amphisiphon on the basis of their con- nate stamens, but no attempt has been made at revising tribal groupings.
* Present address: State Herbarium, Botanic Garden, Adelaide, S. Australia. Accepted for publication 3rd January, 1975.
67
68 Journal of South African Botany
Generic concepts in the Scilleae still require attention. Jessop (1970 & 1972) has proposed a few generic changes, but there are several genera which may require redefinition—for example Massonia and Drimia. The genera at present recognised (excluding certain species belonging to the Aloineae) are:— Schizo- basis, Bowiea, Urginea, Rhadamanthus, Thuranthos, Drimia, Litanthus, Mas- sonia, Daubenya, Androsiphon, Amphisiphon, Neobakeria, Polyxena, Hyacin- thus, Veltheimia, Eucomis, Whiteheadia, Lachenalia, Ledebouria, Scilla, Drimiop- sis, Neopatersonia, Galtonia, Pseudogaltonia, Ornithogalum, Elsiea, Albuca and Dipcadi.
In an attempt to find a meaningful grouping of these genera, certain ap- proaches have been tried, including the use of scanning electron microscope photographs of seed surfaces.
The assistance of Mr. R. H. M. Cross (Officer-in-charge of the Electron Microscopy Unit, Rhodes University) is gratefully acknowledged.
MATERIALS AND METHODS
The seed of 49 plants belonging to 43 species in 25 genera of Liliaceae were examined. Most of the material used was taken from herbarium specimens, but some fresh material was used. No differences in the seeds appeared to result from the use of dried material.
A thin layer of gold palladium alloy was evaporated onto the surface of the samples. The material was examined under a J.E.O.L. JSM U3 scanning electron microscope and in all cases photographs were taken.
OBSERVATIONS
The following brief descriptions are based on the photographs. Because of the limited material examined it is not considered justified to draw up more detailed descriptions.
Bulbine
B. frutescens (Ex hort., Rhodes University)
The concave, more or less isodiametric, cells with several papillae are not similar to those in any other material examined. Cells measure 160-240 um diameter (Fig. 1A). This plant was included for comparison but is not bulbous.
Schizobasis S. intricata (Klaasvoogds, Robertson (3319D), Esterhuysen 22711, BOL). Sub-rectangular ridges probably represent cell margins, but their irregular arrangement suggests that they may be wrinkling, perhaps as a result of drying.
Studies in the Bulbous Liliaceae in South Africa 69
Fie. I Scanning electron microscope photographs of Liliaceae seed surfaces. A. Bulbine frutescens (Ex hort., Rhodes University) x 300. B. Schizobasis intricata (Esterhuysen 22711) x 700. C. Bowiea volubilis (Galpin s.n.) x 200. D. Albuca patersonii. (Martin s.n.) x 700. E. Galtonia princeps (Strey 4516) x 200. F. Drimia anomala (Esterhuysen 18343) x 700.
70 Journal of South African Botany
There is some similarity with several species of Drimia and Urginea. ‘Cells’ concave with smooth surfaces (Fig. 1B).
Bowiea B. volubilis (Byrne, Richmond, Natal (2930C), Galpin s.n., BOL). Rectangular convex cells with slightly undulate margins and a smooth surface are characteristic, but it is somewhat intermediate in form between Urginea modesta and U. marginata. It is also similar to Galtonia princeps. Cells measure 120-130 x 40-50 pm (Fig. 1C).
Albuca
A. canadensis (Huguenot College, Wellington, without collector, sub RUH 4473, RUH).
Cells with smooth surface and undulate margins, 130-180 x 20-30 um.
A. patersonii (Without locality, Martin s.n., sub RUH 9548, RUH).
Cells with smooth surface and more deeply undulate margins than in A. canadensis, c. 90-120 xc. 30 wm (Fig. 1D).
These two species are very similar to one another. The cell outline is similar to that of some species of Urginea and/or Ornithogalum but in those genera there are conspicuously raised areas in the seed surface which are absent in A/buca. There is a strong similarity with Neopatersonia uitenhagensis, which has pits however, and with Galtonia princeps, which has more rectangularly shaped cells.
Galtonia
G. princeps (Port Edward 3130A), Strey 4516, PRE).
Cells with smooth surface and undulate margins, c. 120-150 xc. 30 pum. The sub-rectangular cells are more deeply undulate than, but similar to, those of Bowiea volubilis (Fig. 1E).
Drimia
D. anomala (Uniondale, near Joubertina (3323D), Esterhuysen 18343, PRE).
Cells with contorted surface, minutely rough (the protuberances less than 1 »m diameter), 5- or 6-angled, 40-60 x 20-30 «xm (Fig. 1F).
D. media (Great Drakenstein (3418B), Esterhuysen 21348, BOL).
Cells with more or less smooth surface, more or less rectangular, 80-200 x 40-50 pm (Fig. 2A).
Because of the probably very close relationship, Drimia is discussed at the end of the description of Urginea.
Studies in the Bulbous Liliaceae in South Africa 71
Scanning electron miscroscope photographs of Liliaceae seed surfaces. A. Drimia media
(Esterhuysen 21348) x 200. B. Urginea delagoensis (Buitendag 118) x 2000. C. Urginea basutica
(Strey 3128) x 200. D. Urginea fragrans (Godfrey VH-1254) x 700. E. Urginea pusilla (Marloth 857) x 700. F. Urginea sanguinea (Marloth 7686 )x 700.
72 Journal of South African Botany
Urginea
U. delagoensis (Nelspruit (2530B), Buitendag 118, PRE).
Cells with pitted surface (the pits up to 1-1,3 wm diameter), otherwise smooth, rectangular, 300-400 40-60 pm (Fig. 2B).
U. basutica (Willow Glen, Pretoria (2528C), Strey 3128, PRE).
Cells with numerous curled threads and ribbons and pits (the pits c. 1-2 um diameter), irregular in outline, convex, 80-230 x 70-120 pm (Fig. 2C). The threads and ribbons are removed by boiling briefly in water and are likely to consist of wax. They are, however, not soluble in Xylol at 20 °C.
U. fragrans (7 miles S. of Klaver (3118D), Godfrey VH-1254, PRE).
Cells with the surface covered by a dense reticulation of minute ridges (the ridges 1—-1,5 wm thick), concave, the margins undulate and indicated by distinct grooves, 230-300 x 30-50 xm (Fig. 2D).
U. pusilla (Bultfontein, Kimberley (2824D), Marloth 857, PRE).
Cells with smooth surface, irregular in outline, 110-240 x 40-70 um (Fig. 2E).
U. sanguinea (Without locality, Marloth 7686, PRE).
Cells with pitted surface (pits 1-1,5 ~m diameter), with deeply undulate margin, 200-300 x 30-50 um (Fig. 2F).
U. modesta (near Premier Mine (2528C), Young s.n. sub Tvl. Mus. 32493, PRE):
Cells with almost smooth surface, rectangular, c. 65 x 20-25 «xm (Fig. 3A).
U. modesta (Albany district above Rabbits Wood (3326B), Jessop 853, RUH).
Cells not distinguishable, with irregular surface (Fig. 3B).
U. multisetosa (Willow Park, Pretoria (2528C), Strey 3054, PRE).
Cells with smooth surface, with undulate margins, 100-130 x 50-60 pm.
U. multisetosa (Duiwelskloof (2330C), Scheepers 686, PRE).
Cells with smooth surface, with irregular margins, 150-210 x40-100 »m (Fig. 3C).
U. marginata (Richtersveld, Pillans 5780, BOL).
Cells with smooth surface, with undulate margins, 120-150 xc. 50 »m (Fig. 3D).
Drimia and Urginea can be shown, on several bulb and inflorescence charac- ters, to be very closely related. Discussion of the seeds of both is, therefore, treated together. There is considerable variation in cell shape, but the cells are generally clearly recognisable. In one specimen of U. modesta they are sub- rectangular, but in the other specimen of the same species they are probably constricted in the centre although the outline is poorly defined. Several speci- mens have rather deeply undulate margins (e.g. U. fragrans). In a few specimens (e.g. U. delagoensis) the boundaries of cells are poorly defined. In surface charac- ters there is also considerable variation from smooth (e.g. U. pusilla) to pitted (e.g. U. delagoensis). In U. fragrans there is a reticulation of minute ridges.
Studies in the Bulbous Liliaceae in South Africa Te
Fic. 3. Scanning electron microscope photographs of Liliaceae seed surfaces. A. Urginea modesta (Young s.n.) x 700. B. Urginea modesta (Jessop 853) x 700. C. Urginea multisetosa (Scheepers 686) x 700. D. Urginea marginata (Pillans 5780) x 700. E. Litanthus pusillus (Jacot Guillarmod 3233) x 700. F. Rhadamanthus platyphyllus (Hall 3473) x 200,
74 Journal of South African Botany
Because of the variation it is not possible to compare these genera as a whole with other genera. There does not appear to be any correlation with the normal division of this group into two genera in the sense of, for example, Phillips (1951).
Litanthus
L. pusillus (Lesotho, Mamathes (2927B), Jacot Guillarmod 3233, RUH).
Cells mostly not distinguishable, with a considerably folded surface, rough with ridges up to 1 um thick, 5- or 6-angled, c. 20-25 um diameter (Fig. 3E).
The rather poorly defined cells are somewhat similar to those of one speci- men of Urginea modesta, but the rugulose surface is unique amongst the samples examined.
Rhadamanthus
R. platyphyllus (Strandfontein, Vredendal (3118C), Hall 3473, NBG).
Cells smooth, 4- to 7-angled, isodiametric to oblong, depressed, 80-240 ym diameter. Underlying cells forming shallow convexities in the surface of the superficial cells (Fig. 3F).
R. sp. (Ex hort., Rhodes University).
Cells smooth, 5- or 6-angled, depressed, 70-150 «50-70 wm. Underlying cells forming shallow convexities in the surface of the superficial cells.
In both specimens examined the cells are exceptionally clearly delimited by prominent marginal ridges. The underlying cells are also distinct. It is possible that the gold palladium may not have been thick enough to prevent a large number of electrons from passing through the surface of the seed. But as this phenomenon was not observed in samples of other genera it is believed to be due to a characteristic in the seed surface.
Dipcadi
D. ciliare (Bushman’s River Mouth (3326D), Noel s.n. sub RUH 12089, RUH).
Cells pitted with numerous shallow pits c. 1 wm diameter, 5- or 6-angled, 60-130 x 30-45 ym. Some cells appeared to be abruptly raised above the sur- rounding cells, but this may be an artefact (Fig. 4A).
D. viride (Burnt Kraal, near Grahamstown (3326B), Martin s.n. sub RUH 9271, RUH).
Cells smooth, 5- to 7-angled, 40-50 zm diameter (Fig. 4B).
In both of the species examined the cells are 5- or 6-angled (rarely 7-angled), not often more than twice as long as broad. The most similar species examined are in the Urginea-Drimia group (in which the cells are not as distinctly angled, however), and in Rhadamanthus (in which the margins are raised and the under- lying cells visible).
Studies in the Bulbous Liliaceae in South Africa ID
ow
Fic. 4.
Scanning electron microscope photographs of Liliaceae seed surfaces. A. Dipcadi ciliare
(Noel s.n.) x 700. B. Dipcadi viride (Martin s.n.) x 700. C. Scilla firmifolia (Ex hort., Rhodes
University) x 700. D. Scilla nervosa (Leendertz s.n.) x 700. E. Scilla natalensis (Mogg 2238) x 700. F. Ledebouria floribunda (Van der Merwe 2240) x 700.
76 Journal of South African Botany
Scilla
S. firmifolia (Ex hort., Rhodes University).
Cells distinctly papillose (the papillae rather flattened with three grooves), 5— or 6—angled, 25-35 um diameter (Fig. 4C).
S. nervosa (Warmbaths (2428C), Leendertz s.n. sub Tvl. Mus. 28064, PRE).
Cells indistinct, probably 5— or 6—-angled, 30-40 »m diameter; structures resembling thin-walled papillae present (Fig. 4D).
S. natalensis (Pietermaritzburg (2930C), Mogg 2238, PRE).
Cells indistinct, probably 5— or 6-angled, 30-40 »m diameter; structures resembling thin-walled papillae present (Fig. 4E).
Three of the four species of Scilla recognised by Jessop (1970) were examined. In both S. nervosa and S. natalensis there is a wrinkling of the surface which may be caused by the collapse of the convex outer walls. In S. firmifolia, how- ever, there are large papillae each with I—3 (usually 3) furrows. If the appearance of the former two species is caused by the collapse of thin-walled papillae, the differences between them and S. firmifolia may not be as great as the first im- pression of the photographs may suggest. Although the cells of S. nervosa and S. natalensis are indistinct it is thought likely that all species have 5— or 6— angled cells. These observations give support to the retention of S. nervosa in Scilla rather than in Schizocarphus as proposed by Van der Merwe (1943).
Ledebouria
L. floribunda (Happy Rest, Soutpansberg, N. Tvl., Van der Merwe 2240, PRE).
Surface very irregular; the outermost layer (wax-like in appearance) ap- parently damaged. Cells arranged in regular rows, more or less 4-angled, 15-25 pm diameter (Fig. 4F).
It is not certain that the cell-like patterns are in fact cells as the larger cell- like shapes are subdivided by shallower grooves. There is a close similarity with the Massonia group and possibly with Lachenalia. The round structure in the photograph is thought to be a result of damage.
Eucomis
E. autumnalis var. clavata (Tabamhlope, Estcourt (2929B), Downing 53A, PRE).
Shallow depressions over whole surface c. 20-40 pm diameter. Minute, more or less reticulate, ridges c. 0,1 wm wide over both raised and depressed portions, otherwise smooth (Fig. 5A).
The surface shows no evidence of cells and is similar to some members of the Massonia group.
Studies in the Bulbous Liliaceae in South Africa 77
Fic. 5.
Scanning electron microscope photographs of Liliaceae seed surfaces. A. Eucomis autumnalis
var. clavata (Downing 53A) x 700. B. Ornithogalum thyrsoides (Bought at Houtbay Neck)
x 700. C. Ornithogalum conicum (Leipoldt 4374) x 700. D. Ornithogalum ceresianum (Goatcher
s.n.) x 700. E. Ornithogalum miniatum (Muir 2988) x 700. F. Pseudogaltonia clavata (Télken s.n.) x 700.
78 Journal of South African Botany
Ornithogalum
O. thyrsoides (Table Mountain above Kirstenbosch (3418A), Bolus 4919, PRE).
Cells with smooth surface but separated from one another by deep grooves, with undulate margins, 110-150 xc. 50 pm.
O. thyrsoides (Bought at Houtbay Nek, PRE).
Cells with smooth surface but separated from one another by deep grooves, with undulate margins, 110-150 xc. 40 wm (Fig. 5B).
O. conicum (Clanwilliam (3218B), Leipoldt 4374, PRE).
Cells papillose (the papillae verrucose), with undulate margins, 90-140 x 30-40 um (Fig. 5C).
O. ceresianum (Ceres (3319A), Goatcher s.n. sub BOL 13726, PRE).
Cells with smooth surface, with undulate margins, 90-130 « 40 »m (Fig. 5D).
O. maculatum (Sandberg near Clanwilliam (3218B), Leighton s.n. sub BOL 2161, PRE).
Cells with minutely verrucose surface, deeply grooved between the cells, with undulate margins, 65 x 20 um.
O. miniatum (Near Riversdale (3421A), Muir 2988, PRE).
Cells with numerous shallow depressions c. 1,0 wm diameter, with undulate margins, c. 70 20 wm. Each cell produced into a papilla up to 70 um long (Fig. SE):
O. miniatum (Kommaggas (2917D), without collector sub Stellenbosch University Garden no. 3983, PRE).
Cells with numerous shallow depressions c. 1,0 »m diameter, with undulate margins, 70-80 xc. 20 wm. Each cell produced into a papilla up to 80 um long.
The seven specimens examined were all very similar, with deep grooves be- tween the cells and the outer tangential walls strongly convex or produced into papillae. There is a similarity with Galtonia princeps and, to a limited extent, with Albuca and the Urginea-Drimia group, but these genera have flatter cells.
Pseudogaltonia
P. clavata (Sturmyeld, Gobabis (2119), Té/lken s.n.).
Cells with surface reticulately patterned with shallow depressions c. 20 wm diameter, sub-isodiametric, 5- to 7-angled, 65—100 zm diameter (Fig. 5F).
The cell pattern is similar to that in Scilla and Dipcadi, but the surface re- ticulation is more similar to that of a few species of the Urginea-Drimia group.
Lachenalia L. monophylla (Storms River (3323D), Seagrief s.n. sub RUH 22228, RUH). Surface wrinkled, with a pattern of subrectangular cells c. 30-40 x 30 pm (Fig. 6A).
Studies in the Bulbous Liliaceae in South Africa 79
Scanning electron microscope photographs of Liliaceae seed surfaces. A. Lachenalia mono-
Phylla (Seagrief s.n.) x 700. B. Whiteheadia bifolia (Hardy 542) x 200. C. Neobakeria angusti-
folia (Friedlander s.n.) x 700. D. Massonia depressa (Marloth 10367) x 700. E. Massonia echinata (Bayliss BS/4868) x 3000. F. Massonia putulata (Acocks 21339) x 700,
80 Journal of South African Botany
The absence of surface details allies this specimen with the Massonia group.
Whiteheadia
W. bifolia (Ratelpoort between Springbok and Steinkopf (2917D), Hardy 542, PRE):
Surface wrinkled, with rows of more or less 4-angled cells 10-23 wm diameter (Fig. 6B).
Examination of freezing microtome sections confirmed that the cell-like convexities agree in size with the epidermal cells and are, therefore likely to indicate the cell walls.
Neobakeria
N. angustifolia (De Aar (3024C), Friedlander s.n., GRA). Surface smooth except for cells which are 5- or 6-angled and 20-30 xc. 10 pm. This seed belongs to the Massonia group (Fig. 6C).
Massonia
M. depressa (Without locality, Marloth 10367, PRE).
Surface smooth except for cells arranged in rows and 7-10 ym diameter (Fig. 6D). Freezing microtome sections confirmed that these patterns do repre- sent cells.
M. echinata (Near Klipplaats Village (3324A), Bayliss BS/4890).
Surface coarsely verrucose with ridges down to less than 1 »m diameter. Cells not distinguishable.
M. echinata (Near the Redhouse turnoff (3325D), Bayliss BS/4868).
Surface coarsely verrucose with papillae and ridges down to less than 1 ~m diameter. Cells not distinguishable (Fig. 6E).
M. pustulata (Locality not recorded, Acocks 21339, PRE).
Surface smooth except for more or less square cells c. 10-15 ~m diameter (Fig. 6F).
Epidermal cells are either not recognisable or (as in Whiteheadia) there are rectangular or triangular cells.
Daubenya
D. aurea (Sutherland (3220A), Marloth 7106, PRE).
Surface smooth except for cells which are 4- to 6-angled and 18-23 pm diameter (Fig. 7A).
The cell pattern is similar to that in Whiteheadia and in some species of Massonia.
Studies in the Bulbous Liliaceae in South Africa 81
IG: Scanning electron microscope photographs of Liliaceae seed surfaces. A. Daubenya aurea (Marloth 7106) x 700. B. Androsiphon capense (Acocks 18606) x 700. C. Polyxena ensifolia (Dyer 99a) x 700. D. Hyacinthus corymbosus (NBG 374/64) x 700. E. Neopatersonia uitenha- gensis (Long 135) x 700. F. Asparagus virgatus (Ex hort., Rhodes University) x 700.
82 Journal of South African Botany
Androsiphon
A. capense (Farm Plaatberg, S.W. of Calvinia (3119B), Acocks 18606, PRE).
Cells 4- to 6-angled, 13-23 wm diameter. Surface naked or densely lepidote; the scales irregular in form and arrangement and up to c.3 ym broad (Fig. 7B).
At low magnifications (100) the surface appears indistinguishable from other members of the Massonia group. However, higher magnifications showed a unique scaly covering of one of the four specimens examined.
Polyxena
P. ensifolia (Oudtshoorn (3322C), Dyer 99a, GRA).
Cells not distinguishable. Surface more or less smooth with shallow pits up to c.l10 wm diameter (Fig. 7C). There is a close similarity with Neobakeria angustifolia and Hyacinthus corymbosus.
Hyacinthus
H. corymbosus (Witklip Farm, Vredenburg, without collector sub NBG 374/64, NBG).
Cells not distinguishable. Surface more or less smooth with shallow pits up to c.10 ~m diameter (Fig. 7D).
There is a close similarity with Neobakeria angustifolia and Polyxena ensi- folia.
Neopatersonia
N. uitenhagensis (Bethelsdorp (3325D), Long 135, GRA).
Cells pitted (the pits c. I1-1,5 mm diameter), with undulate margins, c. 130-170 x 60-70 um (Fig. 7E).
The shape of the cells is similar to that of A/buca and to some specimens of the Urginea-Drimia group. The pits are reminiscent of the Urginea-Drimia group or of Pseudogaltonia.
Asparagus
A. virgatus (Ex hort., Rhodes University).
Cells smooth, slightly concave, separated from one another by grooves with almost vertical sides, c. 5-6 wm deep, isodiametric, 5- or 6-angled, 30-40 um diameter. The raised tubular cells are unlike any others in the material examined (Fig. 7F). Asparagus was included for comparison but is not bulbous.
SEED-SURFACE TYPES
Several essentially different types of seed surfaces were observed in the bulbous Liliaceae.
Studies in the Bulbous Liliaceae in South Africa 83
1. The Massonia group (Whiteheadia, Lachenalia, Ledebouria, Massonia, Neo- bakeria, Eucomis, Daubenya, Androsiphon, Polyxena and Hyacinthus) possess either no indication of cells or rectangular or triangular shapes (generally about 15 »m diameter) cells.
2. The Ornithogalum type (Ornithogalum) possesses deeply separated cells with strongly undulate margins.
3. Cells with deeply undulate margins, but flatter outer tangential walls than in Ornithogalum (Albuca, Galtonia and Neopatersonia).
4, Angular cells (Scilla, Dipcadi, Rhadamanthus and Pseudogaltonia).
5. Drimia, Urginea, Litanthus, Schizobasis and Bowiea do not fit exactly into any of the above types although certain species of the Urginea-Drimia group could be allocated to types 3 or 4. Other species have less deeply undulate out- lines than in the previous types or more or less rectangular cells.
GENERIC GROUPINGS
The Bowieae (sensu Hutchinson) deviate further from the norm of bulbous Liliaceae than do any of the other previously proposed groups. Within this tribe there are, however, two very different elements. Bowiea and Schizobasis have bulbs, no foliage leaves in the mature plant and spurred bracts. Eriospermum has a tuber, foliage leaves (usually hysteranthous) and bracts without spurs. Eriospermum, in the absence of a bulb, is not discussed further here.
The affinities of Bowiea and Schizobasis are of interest. They differ from any other South African members of the bulbous Liliaceae in lacking foliage leaves in the mature plant (although they are present in juvenile stages), and in having a branched aerial stem (present also in Scilla firmifolia). These two genera resemble several other genera in, what are considered by the present author to be, a number of important characters. These characters are the spurred bracts, absence of starch from the bulb, and loose testa. The genera to which these characters ally them are Drimia, Urginea, Rhademanthus, Thuranthos and Litanthus. The flowers and bulbs are similar and the present work on seed surface characters tends to support this conclusion. A basic chromosome number of 10 is probably also a common feature, although further data are required.
Two evolutionary processes must be postulated to explain the relationship between these two groups:
(a) reduction in the size of the foliage leaves which are present in the seedlings of Bowiea and Schizobasis and at all stages of the other genera, but in the former two genera are reduced, to bulb scales only, in the mature plant;
(b) development of branches in the inflorescence.
It is suggested that these evolutionary developments, while of undoubted systematic importance, do not justify subdivision of the proposed group in-
84 Journal of South African Botany
cluding Bowiea, Schizobasis, Urginea, Drimia, Rhadamanthus, Thuranthos and
Litanthus. So far as is known there are no genera belonging to this group not
represented in the South African flora. A revision of Rhadamanthus has recently
been published (Nordenstam, 1970) and a revision of the remaining genera will be published shortly by the present author.
The only other tribe split off the Scilleae has been the Massonieae (Hutchin- son, 1959). Seed surface characters, as shown in this paper, appear to link several genera, including the two placed in the Massonieae (Massonia and Daubenya). These genera (Massonia, Daubenya, Whiteheadia, Lachenalia, Ledebouria, Neobakeria. Eucomis, Androsiphon, Polyxena and Hyacinthus), and Amphisiphon and Drimiopsis which were not studied, have morphologically similar bulbs, leaves and inflorescences. Most have representatives with leaves marked with various shades of green or red-purple. Neopatersonia was regarded by Phillip (1951) as also belonging to this group but seed surface characters appear to provide strong evidence for separating it. Although there are no strong characters for uniting the members of this group, except for the seed surface characters, there are several features uniting various combinations of genera within the group. These are:—
(a) Lachenalia, Ledebouria and Drimiopsis appear to have axillary inflorescences (Jessop, 1972). The rest of the genera have not been determined.
(b) Lachenalia pusilla, which has very slightly zygomorphic flowers, links Lachenalia (zygomorphic) with Polyxena (actinomorphic). Both genera and Hyacinthus (Cape species only) have a tubular membranous bulb scale as well as fleshy scales.
(c) The Eucomis flower is similar to that of Massonia, and Whiteheadia forms an intermediate between these genera. The coma occurs in one species of Neobakeria as well as in Eucomis, although it is not as well developed in the Neobakeria. A coma does also occur in a few genera not included in this group—for example in Muscari (European and Asian).
(d) Drimiopsis and Ledebouria are separated only on the incurving, subcu- cullate perianth of the former. In all other respects these two genera are extremely close.
(e) Massonia forms a natural group with certain other genera as has been pointed out by Phillips (1951). The members of this group are separated from one another on a fairly small number of characters—for example Daubenya from Massonia on the zygomorphic flower, Neobakeria from Massonia on the absence of large spathe-like bracts, Polyxena from Neo-. bakeria on the insertion of the filaments and the membranous bulb scales.
The remaining genera, not placed in the Urginea-Drimia group or the Massonia group, from a heterogenous assemblage, with little to link the genera. Scilla, Ornithogalum and Dipcadi occur in South Africa and in Europe and Asia.
Studies in the Bulbous Liliaceae in South Africa 85
Galtonia, Pseudogaltonia, Albuca and Neopatersonia are principally South African but endemic to Africa. Camassia occurs in North America only, Chionodoxa, Hyacinthus (excluding the Cape species belonging to the Massonia group), Muscari and Puschkinia are northern hemisphere in distribution, mainly from Europe and Asia but also North Africa. Rhodocoedon is endemic to Mada- gascar. Because of the wide distribution of this group the material has not been available to the present author to examine fully the affinities of these genera. The following grouping of those genera (other than members of the Aloineae), which have bulbs and which occur in South Africa (or are closely allied to South African genera) is proposed :— 1. Tribe Scilleae sensu Hutchinson (1959), emend. Bulb containing starch, leaves not spotted or marked, bracts not spurred, seeds with tightly fitting testa and a distinct cellular pattern superficially. Scilla, Ornithogalum, Dipcadi, Galtonia, Pseudogaltonia, Albuca, Camassia, Chionodoxa, Rhodocodon, Hyacinthus (northern hemisphere spp), Muscari, Puschkinia and Neopatersonia. 2. Tribe Massonieae sensu Hutchinson (1959), emend. Bulb containing starch, leaves usually spotted or marked, bracts not spurred seeds with tightly fitting testa and indistinct or no cellular pattern super- ficially. Massonia, Neobakeria, Polyxena, Hyacinthus (Cape spp), Daubenya, Andro- siphon, Amphisiphon. Ledebouria, Drimiopsis, Lachenalia, Veltheimia, Eucomis. 3. Tribe Bowieae sensu Hutchinson (1959), emend. Bulb not containing starch, leaves not spotted or marked, bracts spurred (or at least the lower), seeds with loosely fitting testa and a distinct cellular pattern superficially. Urginea, Drimia, Rhadamanthus, Litanthus, Thuranthos, Schizobasis and Bowiea.
REFERENCES
Hutcuinson, J., 1959. The families of flowering plants 2. 2nd edition. Oxford: Clarendon Press.
Jessop, J. P., 1970. Studies in the bulbous Liliaceae: 1. Scilla, Schizocarphus and Ledebouria. JIS. Afr. Bot. 36 (4): 233-266.
Jessop, J. P., 1972. Studies in the bulbous Liliaceae in South Africa: 2. Drimiopsis and Resnova. JIS. Afr. Bot. 38 (3): 151-162.
Krause, K., 1930. Liliaceae. In: A. Engler, (ed.) Die natiirlichen Pflanzenfamilien. 2nd edition. Leipzig: W. Engelmann.
NorDENSTAM, B., 1970. Studies in South African Liliaceae: 3. The genus Rhadamanthus. Bot. Notiser 123 (1): 155-182.
Puiutrs, E. P., 1951. The genera of South African flowering plants. 2nd edition. Mem. bot. Surv. S. Afr. 25.
VAN DER MERWE, F. Z., 1943. Schizocarphus nervosus. Flower. Pl. S. Afr. 23: t.904.
JIS. Afr. Bot. 41 (2): 87-90 (1975) a
DIE KARIOTIPE VAN OROTHAMNUS ZEYHERI (PROTEACEAE)
K YOICHI OGAWA (The Institute for Breeding Research, Tokyo University of Agriculture)
PHILLIP VAN DER MERWE*
(Seksie Natuurbewaring, Provinsiale Administrasie van die Provinsie die Kaap die Goeie Hoop)
UITTREKSEL
Wortelpunte van twee kiemplante van Orothamnus zeyheri Pappe is sitologies ondersoek en die chromosoomgetal vasgestel op 2n = 24
ABSTRACT
THE CARYOTYPE OF OROTHAMNUS ZEYHERI (PROTEACEAE) Chromosome counts from smears of root tips of two seedlings of Orothamnus zeyheri Pappe yielded 2n = 24 chromosomes.
INLEIDING
De Vos (1943) het die chromosoomgetalle van agt genera van die Suid-Afri- kaanse Proteaceae vasgestel.
Venkata Rao (1971) tabelleer die gegewens van De Vos (1943), Lancaster (1952), Ramsay (1963), Johnson en Briggs (1963), Venkata Rao (1957) en Hair en Beuzenberg (1959).
Vir *n plant soos Orothamnus wat met uitsterwing bedreig word, is dit belangrik om die chromosoomgetal te bepaal ten einde die naaste verwante te vind vir moontlike onderstokke vir ent- en okulasie-doeleindes.
Gedurende Julie 1974 is ’n poging aangewend om die haploiede chromo- soomgetal te bepaal van verdelende generatiewe kerne in die stuifmeelbuise van Orothamnus-stuifmeel vanaf die Eddie Rubenstein Orothamnus Reservaat by Hermanus, Kaapprovinsie, Republiek van Suid-Afrika. Chromosome is wel waargeneem, maar nie duidelik en genoegsaam uitgesprei om uitsluitsel te kan gee nie. ’n Telling van dieselfde tydelik gemonteerde preparaat het ’n haploiede chromosoomgetal gelewer van n = 10(P. van der Merwe) enn = 12 (D. van der Merwe).
* Preliminére resultate vir die Ph.D.-graad aan die Universiteit van Stellenbosch, onder promotorskap van Prof. Miriam P. de Vos. Vir publikasie aanvaar 28 Februarie 1975.
87
88 Journal of South African Botany
MATERIAAL EN METODES
Orothamnus-vrugte wat versamel was op 30/11/73 in die Eddie Rubenstein- Reservaat is op 31/12/73 gesaai by die Assegaaiboschkwekery, Jonkershoek, Stellenbosch en het ontkiem gedurende Augustus 1974. Een van hierdie kiem- plante het verrot en omgeval nog voordat die epikotiel begin verleng het. Die kiemwortel en die voet van die hipokotiel was totaal deur die infeksie vernietig. Die verrotte deel is op 15/10/74 met water weggewas en die gesonde deel van die hipokotiel met 0,1°% 4(3-indoliel)bottersuur behandel en in ’n nie-steriele grondmengsel geplant. Op 28 November 1974 om 11.0 v.m. is hierdie kiemplant, waarvan die epikotiel intussen verleng het, uitgehaal en is gevind dat wortel- vorming aan die basis van die hipokotiel en vanuit die stele-gebied, plaasgevind het. Die wortels is in lopende water gewas en die wortelpunte afgesny en in ’n oplossing van 0,001M 8-hidroksikinolien in water geplaas vir 3 uur en ge- nommer 247 (P.v.d.M.). Hierna is die materiaal gefikseer in Asynalkohol (3 Etielalkohol: 1 Ysasyn) en gekleur in 45°, Asynkarmyn vir 24 uur waarna platgeperste mikroskooppreparate gemaak is. ’n Tweede sterk kiemplant (5 cm) is ook ontwortel en die wortelpunte soortgelyk behandel en genommer 248. Die twee kiemplante is met 0,2°% bottersuur behandel en weer geplant en onder- skeidelik genommer 247 en 248 vir verwysingseksemplare.
Fic. 1
Die Kariotipe van Orothamnus zeyheri (Proteaceae) 89
a ad o " e @ ¢ B fa an [ea] < ig ° ° ? e ry a a 1Q@ 4m
RESULTATE
Die tydelike platgeperste preparate in 45°/, asynsuur toon 2n = 24.
Die chromosome toon effense verskille in lengte en is ongeveer 2 wm lank in hul effens verkorte toestand, te wyte aan die voorbehandeling met 8-hidrok- sikinolien. Hulle sentromere is mediaan of sub-mediaan en twee van die chromo- some het elk ’n klein sateliet (0,5 wm lank). Hulle vertoon heelwat korter as die wat Horn (1962) vir Protea latifolia aantoon. Die kortste van P. Jatifolia se chromosome word aangedui as 2,2 wm terwyl dit ongeveer die lengte van die langste van Orothamnus se chromosome is. Hierdie resultate is egter nie abso- luut vergelykbaar nie aangesien Horn as voorbehandeling ’n versadigde oplos- sing van monobromonaphtaleen gebruik het waarvan die krimpingseffek nie noodwendig dieselfde is as 8-hidroksikinolien nie. Die stadium wat Horn ondersoek het, was ’n laat profase, terwyl hierdie ’n metafase-stadium was.
GEVOLGTREKKINGS
Die telling n = 12 deur D. van der Merwe word bevestig deur K. Ogawa se telling van 2n = 24. Hierdie telling van 2n = 24 vir Orothamnus is in ooreen- stemming met die resultate van De Vos (1943) wat n = 12 by vyf van die twaalf
90 Journal of South African Botany
genera van die tribus Proteeae gevind het, naamlik Protea, Leucospermum, Mimetes, Paranomus en Serruria. Slegs die tweehuisige Leucadendron was n = 13 (wat ooreenstem met die vier Australiese genera van hierdie tribus). Sover bekend, kom n = 12 nie by die Australiese genera van die Proteaceae voor nie.
Die ander genera van die Suid-Afrikaanse Proteaceae wat Venkata Rao (1971) in aparte tribusse plaas, naamlik Aulax (tribus Aulaxeae) en Brabeium (tribus Macadamieae) se chromosoomgetalle is onderskedelik n = 11 enn = 14 (De Vos, 1943).
Vegetatiewe verenigbaarheid tussen Orothamnus en Leucospermum, Protea, Mimetes, Paranomus en Serruria (almal n = 12, de Vos, 1943) behoort dus uitgetoets te word.
DANKBETUIGING Mnr. G. C. Crafford was verantwoordelik vir die fotografie.
LITERATUURVERWYSING
De Vos, M. P., 1943. Cytological studies in genera of the Proteaceae. S. Afr. J. Sc. 40: 113-122. Horn, W., 1962. Caryotype analysis in Protea L. JI] S. Afr. Bot. 28: 255-257.
VENKATA RAO, C., 1971. Proteaceae. (Botanical monograph no. 6.) New Delhi: C.S.I.R. of India.
JIS. Afr. Bot. 41 (2): 91 (1975)
’n NUWE NAAM VIR DUTHIELLA
M. P. DE Vos (Universiteit van Stellenbosch)
Dit het onder my aandag gekom dat die naam Duthiella De Vos vir ’n nuwe monotipiese genus van die Iridaceae ’n latere homoniem is van die reeds gepub- liseerde geldige genusnaam, Duthiella C. Muell. ex Broth., in Nat. Pfl. 1: 1009 (1908), vir *n blaarmosgenus behorende tot die familie Trachypodaceae. Die naam Duthiella vir die genus van die Iridaceae is daarom illegitiem en moet verwerp word.
Die ongelukkige fout, wat die reeds ingewikkelde plantkundige nomenklatuur nog meer ingewikkeld maak, laat hierdie genus sonder geldige naam. ’n Nuwe naam moet daarvoor gepubliseer word, gebaseer op dieselfde tipe as dié van die verwerpte naam. My keuse is:
Duthiastrum De Vos, nom. nov. Sinoniem: Duthiella De Vos, nom. illeg., in Jl S. Afr. Bot. 40: 301 (1974): non Broth. 1908.
Die naam van die tipe word dan Duthiastrum linifolium (Phill.) De Vos vir Duthiella linifolia (Phill.) De Vos (Syringodea linifolia Phill. in Ann. S. Afr. Mus. 9: 125 (1913).
Net soos die latere homoniem Duthiella, is die nuwe naam ter ere van dr. A. V. Duthie (1881-1963), die eerste dosent in Botanie aan die Victoria-Kollege, later die Universiteit van Stellenbosch, en ook weens die stervormig spreidende blomdeklobbe van die plant.
i i ET a Vir publikasie aanvaar 4 April 1975.
91
JIS. Afr. Bot. 41 (2): 93-124 (1975)
NEW TAXA AND NEW COMBINATIONS IN THE GENUS CRASSULA
H. R. TOLKEN (Bolus Herbarium, University of Cape Town)*
ABSTRACT
84 new taxa and new combinations are proposed. Keys are provided to some species complexes.
UITTREKSEL
NUWE TAKSA EN KOMBINASIES IN DIE GENUS CRASSULA 84 nuwe taksa en nuwe kombinasies word voorgestel. Sleutels vir enige spesies komplekse is ingesluit.
The need for accepted names in several publications has prompted the author to validate new combinations and to describe new taxa before the complete taxonomic revision of the genus Crassula in Southern Africa appears in print. Author abbreviations as recommended for the Flora of Southern Africa are adopted, so that ‘Schonl.’ is used here for Sch6nland.
_ 1. Crassula alba Forsk. var. parvisepala (Schonl.) Toelken, comb. nov. C. rubicunda var. parvisepala Schonl. in Trans. Roy. Soc. S. Afr. 17: 231 (1929). Type: Transvaal, Lydenburg, Wilms 538 (GRA, lecto!; BM!; E!; G!; P!; Z!). Although the flowers of this species are usually red in South Africa they do not differ morphologically from those of C. alba from northern Africa. Wilms 538 is chosen as lectotype from the three syntypes, as it is well represented in many herbaria.
2. Crassula alpestris Thunb. subsp. massonii (Britt. & Bak. f.) Toelken, comb. nov. et stat. nov.
C. massonii Britt. & Bak. f. in J. Bot., Lond. 35: 485 (1897). Type: Promontorium Bonae Spei, Masson s.n. (BM, holo!; PRE, photo!).
This subspecies, which occurs in two widely separated localities, seems to hybridize with the typical subspecies wherever the two taxa occur in proximity to one another. The subsp. alpestris usually has a terminal thyrsoid inflorescence while the subsp. massonii has terminal and lateral cymules.
LN ETE Er *Present address: Liaison Officer of the Botanical Research Institute, Pretoria at Royal Botanic Gardens, Kew.
Accepted for publication 12th March, 1975.
93
94 Journal of South African ee
3. Crassula ammophila Toelken, sp. nov. ab ie “solhaaltnlla mile adpressis foliorum usque ad | mm longis et marginibus apicum petalorum non recurvis differt.
Frutices perennes caulibus fragilibus erectis usque ad 80 cm longis et juvenibus rare latioribus 0,2 cm in diametro, paucis ramosis, foliis veteribus caducis. Folia oblongo-lanceolates, 0,8-1,5 cm longa, 0,3-0,4 cm lata, obtusa vel acuta, plana vel leviter convexa supra, covexissima subtus, dense tecta pilis recurvis adpressis, cano-viridia vel cano-fusca; vaginae foliorum usque ad 1 mm longae. Hydathodi dispersi supra et subtus, inconspicul. Jnflorescentia thyrsus terminalis oblongatus vel globosus, 1-3 dichasiis caulibus plus minusve longis, floribus sessilibus pentameris, pedunculo 1-3 cm longo et pilis recurvis adpressis, bracteis triangularibus in dichasiis. Sepala triangularia, 2—-2,5 mm longa, obtusa vel acuta, dense tecta pilis expansis et cillis marginalibus saepe indistinctis, succulenta cano-viridia. Petala panduriformia, 3—3,5 mm longa, apicibus acuminatis canaliculatis et cristis dorsalibus ad apices, 0,3-0,6 mm connata, apicibus laborum vix recurvis, eburnea. Stamina c. 2 mm longa, antheris nigris 0,7-0,9 mm longis, filamentis leviter latescentibus ad bases et aliquantum constrictis ubicumque connatis tubo petalorum. Squamae oblongae, 0,7-0,8 x 0,2-0,4 mm, leviter emarginatae, vix constrictae ad basim, parum succulentae, luteae. Carpella ovariis reniformibus constrictibus ad stigmates laterales rubra; ovarium 10-16 ovulis elongatis et papillis inconspicuis serialibus.
Type: Cape, Nuwerus, Pearson 5501 (GRA, holo!; K!).
Perennials with brittle erect branches up to 80 cm long, often scrambling or rarely decumbent shrublets and with young branches rarely more than 0,2 cm in diameter and with peeling bark towards the base, little branched and with old leaves deciduous. Leaves sessile, oblong-lanceolate, 0,8-1,5 cm long, 0,3—-0,4 cm broad, with bluntly acute or rounded apices, flat or slightly convex above and very much convex below and almost as thick as broad, densely covered with recurved adpressed hairs, grey-green to greyish-brown; sheath up to 0,1 cm long and formed by the partial fusion of the leaf bases. Hydathodes scattered mainly over the upper but also on the lower surface, indistinct. Inflorescence a terminal elongated to rounded or globular thyrse with one to three dichasia more or less stalked, with sessile 5-merous flowers, with a peduncle hardly distinct from a vegetative branch and 1-3 cm long and with adpressed hairs, with triangular bracts with rounded apex rather like the sepals. Sepals triangular 22,5 mm long, bluntly acute covered with long erect hairs and with often indistinct marginal cilia, fleshy, green. Petals panduriform, 3—-3,5 mm long, with acuminate canaliculate apices without dorsal ridges, fused into a tube 0,3-0,6 mm long, with apices of the lobes hardly recurved, cream. Stamens c. 2 mm long, with black anthers 0,7-0,9 mm long, with filaments slightly
New Taxa and New Combinations in the Genus Crassula 95
broadened towards the base and somewhat constricted where fused to the petal tube. Squamae oblong 0,7—-0,8 x 0,2-0,4 mm, slightly emarginate, hardly constricted towards the base, fleshy, yellow. Carpels with reniform ovaries gradually constricted towards the sessile and somewhat laterally placed red stigmas; ovary slightly papillose and with 10-16 elongate ovules covered with rows of inconspicuous papillae.
A scrambler among shrubs of the coastal sandveld vegetation from near Saldanha Bay to north of the Holgat River. CAPE—2816 (Oranjemund): Holgat River (-DD), Té/ken 5270 (PRE). —3017 (Hondeklipbaai): 18 km east of Hondeklipbaai (-AD), Wisura 891 (NBG).
—3118 (Vanrhynsdorp): Nuwerus (-AB), Pearson 5501 (BOL, K): near Koekenaap (-AD), Télken 5163 (PRE). ' —3317 (Saldanha): Danger Point (-BB), Hall in NBG 126/53 (BOL, NBG).
4. Crassula arborescens (Mill.) Willd. subsp. undulatifolia Toelken, subsp. nov. ab subsp. arboresenti foliis ellipticis et undulatis differt.
Folia elliptica vel elliptico-oblanceolata, 0,8-1,5 (-2) cm lata, undulata, leviter succulenta et erecta.
¢ Type: Cape, 1 km north of Sapkamma Station, Té/ken 4290 (BOL, holo!).
Leaves elliptic to elliptic-oblanceolate, 0,8-1,5 (—2) cm broad, undulate, slightly fleshy, erect.
On rocky slopes in sheltered kloofs in the southern parts of the Klein Winterhoek mountains.
This subspecies flowers in December and January as is typical for C. arborescens, while the very similar C. ovata, which occurs within a few kilo- metres of the type locality of the new subspecies, flowers during the period June to August.
CAPE—3325 (Port Elizabeth): 1 km north of Sapkamma Station (-AC), Télken 4290 (BOL); Paardepoort, near Glenconner Station (-AC), Télken 4299 (BOL).
5. Crassula atropurpurea (Haw.) Dietr., Syn. Pl. 2: 1031 (1840). Iconotype: plate 780 (K, lecto!).
Globulea atropurpurea Haw. in Phil. Mag. 64: 189 (1824); DC., Prodr. 3: 391 (1828).
This variable complex is viewed as one species with six divergent varieties, each of which occupies a somewhat different ecological niche but their ecological tolerances allow them to occur sufficiently close to one another with the result that extensive hybrid populations are found in some areas,
96 Journal of South African Botany
1 Leaves glabrous: Lateral branches with internodes more or less equal in length; peduncle with (2) 3-5 bracts without part-inflorescences . C. atropurpurea var. atropurpurea 2 Lateral branches with first internodes much longer than subsequent ones; peduncle with (0) 1 (2) pairs of bracts without part-inflorescences . Cc. atropurpurea var. cultriformis I Leaves hairy:
3 Leaves triangular to almost round in section . . C. atropurpurea var. purcellii 3 Leaves dorsiventrally compressed: _ oe Leaves with recurved marginal cilia . . C. atropurpurea var. rubella
4 Leaves without marginal cilia or if present then spreading:
5 Petals with membranous apex curved downwards and with dorsal appendage in terminal position and more or less canali- culate . . . (C. atropurpurea var. anomala
5 Petals without membranous apex and with terminal canali- culate fleshy rostrum:
6 Lateral branches with first internodes much longer than subsequent ones; petals glabrous . . ©. atropurpurea var. cultriformis 6 Lateral branches with internodes more or less equal in length; petals minutely papillose .. C. atropurpurea var. watermeyeri
var. anomala (Schonl. & Bak. f.) Toelken, comb. nov. et stat. nov. C. anomala Schont. & Bak. f. in J. Bot., Lond. 36: 370 (1898); Schonl. in Trans.
Roy. Soc. S. Afr. 17: 263 (1929). Type: Cape, French Hoek, Schlechter 9317 (GRA, holo!; BM!; BOL!; E!; G!; K!; P!; Z!).
var. cultriformis (Friedr.) Toelken, comb. nov. et stat. nov.
C. cultriformis Friedr. in Mitt. Bot. Staatssamml. Miinchen 6: 638, fig. 9A (1967) et in Prodr. Fl. S. W. Afr. 52: 27 (1968). Type: Cape, Holgat River, Herre 4088 (M, holo).
— var. purcellii (Schonl.) Toelken, comb. noy. et stat. nov. C. purcellii Schonl. in Rec. Albany Mus. 2: 148 (1907) et in Trans. Roy. Soc. S. Afr. 17: 266 (1929). Type: Cape, Matjesfontein, Purcell s.n. (GRA, holo!).
var. rubella (Compton), Toelken, comb. nov. et stat. nov. C. rubella Compton in Trans. Roy. Soc. S. Afr. 19: 291 (1931). Syntypes: Cape, Witteberg, Compton 2820 (BOL!); Whitehill, Compton 2892 (BOL!).
—— var. watermeyeri (Compton) Toelken, comb. nov. et stat. nov. C. watermeyeri Compton in J. Bot., Lond. 69: 49 (1931). Type: Cape, near Nieuwoudtville, pecs 3208 (BOL, holo!).
6. Crassula ausensis P. C. Hutch. subsp: giessii (Friedr.) Toelken, comb. nov. et stat. nov.
C. giessii Friedr. in Mitt. Bot. Staatssamml. 11: 330, fig. 5 (1974). Type: South West Africa, Naukluft, Merxmiiller & Giess 28 956 (M, holo). The very short pedicels and usually ciliate leaves distinguish this subspecies from the typical one.
New Taxa and New Combinations in the Genus Crassula 97
7. Crassula barbata Thunb. subsp. broomii (Schonl.) Toelken, comb. nov. et ly stat. nov.
C. broomii Schonl. in Rec. Albany Mus. 2: 146 (1907). Type: Cape, Pampoen- poort, Broom s.n. (GRA, holo!).
The leaves are broadly obovate and these have marginal cilia only up to 0,1 cm long. The very large lateral stigma and the more or less distinct style found in the type specimen are variable characters in this subspecies and seem to depend on the stage of maturity of the carpels.
8. Crassula brachystachya Toelken, sp. nov. ab C. obovata foliis acutis et brevibus qui abrupte abbreviatis ad basim inflorescentiae, floribus sessilibus et appendice brevi in petalo differt.
Suffrutices perennes plerumque aliquot rosulis foliis spiraliter dispositis, caulibus succulentis et fragilibus 10-20 cm longis ubi florentibus et saepe radicantibus crassis sed non tuberosis. Folia sessilia, oblanceolato-elliptica, 2-4 (-8) cm longa, 0,5—1,5 (—2) cm lata, acuta gradatim constricta ad basim, dorso- ventraliter complanata sed aliquantum succulentia, glabra sed ciliis sub- distantibus, viridia vel saepe margine rubrotincto; vaginae foliorum 0,2-0,5 cm longa. Hydathodes in series una densa secus marginem superum et inferum et aliquot dispersae supra, vix visibles. Inflorescentia terminalis spiciformibus, floribus aliquot pentameris in dischasiis sessilibus, caulibus erectis glabris et bracteis folia similis. Sepala oblongo-elliptica, 3,5—-4,5 mm longa, arista brevi in quoque apice acuto, glabra sed ciliis marginalibus, succulentia, viridia. Petala oblongo-ovata, 5-6 mm longa, acuta et appendice dorsali distincta in quoque petalo, connata et tubo c. | mm longa, apicibus loborum recurvis, alba. Stamina 4-5 mm longa, antheris nigris 0,8-1 mm longis, filamentis latis factis ad basim. Squamae transverse oblongae, 0,2-0,3 x 0,8-1 mm, plerumque truncatae, vix constrictae ad basim, leviter succulentae, flavae. Carpella ovariis gracilibus reniformibus, stylis tenuibus et stigmatibus terminalibus parvis; ovarium laevis et 26-34 ovulis elongatis et papillis serialibus.
Type: Cape, Wittekloof, Compton 2819 (BOL, holo!).
Perennials with usually several rosettes with leaves spirally but loosely arranged at the end of fleshy and brittle branches 10-20 cm long when flowering and often with thick but not tuberous roots, with old leaves usually soon wearing off. Leaves sessile, oblanceolate-elliptic, 2-4 (-8) cm long, 0,5—1,5 (-2) cm broad, acute and gradually tapering towards the base, abruptly shortened on the erect branch which ends in the inflorescence, dorsiventrally flattened but somewhat convex, glabrous except for subdistant cilia which are rarely slightly recurved, green or often tinged red along the margin; sheath 0,2-0,5 cm long and formed by the partial fusion of the leaf bases. Hydathodes arranged
98 Journal of South African Botany
in a dense row along the upper and lower margin and few scattered over the upper surface, hardly visible. Inflorescence terminal, spike-like, with several 5-merous flowers in sessile dichasia, with a glabrous erect stem not distinguish- able from a vegetative branch but with short leaf-like bracts. Sepals oblong- elliptic, 3,5-4,5 mm long, with acute apices often with a short awn, glabrous except for some marginal cilia, fleshy green. Petals oblong-obovate, 5-6 mm long, acute and each with a distinct dorsal appendage, fused into a tube c. 1 mm long, with apices of the lobes recurved, white. Stamens 4-5 mm long, with black anthers, 0,8-1 mm long and longer than broad, with filaments somewhat broadened towards the base and constricted where fused to the petal tube. Squamae transversely oblong, 0,2—-0,3 0,8—1 mm, usually truncate, somewhat constricted towards the base, slightly fleshy, yellow. Carpels with slender reniform ovaries and with long thin styles half to two-thirds of the length of the ovaries and with small terminal stigmas; ovary smooth and with 26-34 elongate ovules covered with rows of papillae.
In moist shaded rock crevices on rock faces in ravines and recorded only from the north-eastern Witteberg and northern Swartberg.
CAPE—3320 (Montagu): Wittekloof (-BD), Compton 2819 (BOL); Pieter Meintjies (-AD), Marloth 9987 (PRE).
—3321 (Ladismith): west of Bosluiskloof (-AD), Té/ken 4106 (BOL).
—3322 (Oudtshoorn): near Prince Albert (-AA), Stokoe s.n. (BOL).
9. Crassula brevifolia Harv.\subsp. psammophila Toelken, subsp. nov. ab subsp. brevifolia caulibus et foliis latis, carpellis laevibus et ovulis 10-12; ab C. macowa- niana apicibus obtusis foliorum et sepalis acutis differt.
Fruticuli usque ad 20 cm alti, caulibus effusis vel decumbentibus, parum ramosi, internodiis inter folia brevissimis et caulibus infra partem foliis (0,3)
0,4-0,5 cm lata, obtusa, patentia, glaucesentia vel saepe fusca. Inflorescentia corymbo-thyrsus, pedunculo 4-8 cm longo, bracteis anguste triangularibus plerumque longioribus 2 mm. Pefala oblonga, crista dorsale sed sine appendice dorsale in quoque petalo, alba suffosa rosea. Antherae atro-purpureae. Carpella laevia et 10-12 ovulis.
Type: Cape, hills along the Kammas River on the road between Port Nolloth and the Grootmis, Télken 4004 (BOL, holo!).
Shrublets up to 20 cm high, with spreading branches rarely decumbent, little branched, with internodes between the leaves not visible and branches below the region with leaves (0,3-) 0,4 cm in diameter. Leaves (2-) 2,5-4 cm long, (0,3—) 0,4-0,5 cm broad, with rounded apex, straight and spreading at about right angles to the stem, dull green or more often brownish-red and without bloom. Inflorescence an almost flat-topped thyrse, with a peduncle 4-8 cm
New Taxa and New Combinations in the Genus Crassula 99
long, with narrowly triangular bracts usually longer than 2 mm and longer than broad. Petals oblong and each with a dorsal ridge but usually without dorsal appendage, white tinged pink. Anthers dark purple. Carpels glabrous and with 10-12 ovules.
CAPE—2816 (Oranjemund): Peacock’s Bay (-DA), Télken 3368 (BOL); Holgat River (-DD), Télken 1855 (BOL).
—2916 (Port Nolloth): along Kammas River (-BB), Tél/ken 3465 (BOL); 4404 (BOL). Springbok ?, Herre in SUG 5879 (GRA).
10. Crassula campestris (Eckl. & Zeyh.) Endl. ex Walp. subsp. pharnaceoides (Fisher & C. A. Mey.) Toelken, comb. nov. et stat. nov.
Crassula pharnaceoides Fischer & C. A. Mey. in Ind. Sem. Hort. Petrop. 8: 56 (1841); Walp., Rep. 2: 254 (1843); Friedr. in Prodr. Fl. S. W. Afr. 52: 33 (1968). Type: Abyssinia, near Adoa, Schimper s.n. (B!; G!; GRA!; P!; S!). Tillaea pharnaceoides (Fisher & C. A. Mey.) Steudel, Nomencl. Bot. ed. 2, 2: 687 (1841); Britt., Fl. Trop. Afr. 2: 387 (1871).
~ The seeds of this subspecies are about twice as long as broad while they
a are scarcely longer than broad in the subsp. campestris.
Both Fischer & C. A. Meyer and Steudel derived their name from a printed label inscribed Tillae pharnaceoides but without description by Hochstetter. Fischer and Meyer described it as a species of Crassula so that Steudel’s refer- ence eight months later can be construed as a new combination, as it is based on the same type material.
11. Crassula capensis (L.) Baill. var. albertiniae (Schonl.) Toelken, comb. nov. et stat. nov.
C. albertiniae Schonl. in Ann. Bolus Herb. 2: 93 (1917) et in Trans. Roy. Soc. S. Afr. 17: 207 (1929). Type: Cape, Onverwacht, Muir 1315 (GRA, holo!).
_—— var. promontorii (Schonl. & Bak. f.) Toelken, comb. nov. et stat. nov.
C. promontorii Schonl. & Bak. f. in J. Bot., Lond. 36: 366 (1898) et in Trans. Roy. Soc. S. Afr. 17: 204 (1927). Type: Cape, Kasteelberg, Wolley Dod 1624 (GRA, holo!; SAM!).
The var. promontorii is distinguished from the typical variety by the absence of a peduncle while the var. albertiniae is characterized by its leaves with entire margin and they are often produced after flowering.
12. Crassula capitella Thunb. subsp. capitella
C. capitella Thunb. in Acta Nova Phys.-Med. Acad. Caes. Leop.-Carol., Nat. Cur. 6: 330, 339 (1778) et Fl. Cap. ed. Schultes 286 (1823); L. f., Suppl. 190 (1781); Ait., Hort. Kew. ed. 1, 1: 394 (1789). Type: Cape, between Cannaland and Olifant’s River, Thunberg in Herb. Thunberg 7732a (UPS, lecto!).
100 Journal of South African Botany
C. spicata Thunb. in Acta Nova Phys.-Med. Acad. Caes. Leop.- Carol., Nat. Cur. 6: 329, 333 (1778) et Fl. Cap. ed. Schultes 284 (1823); L. f., Suppl. 189 (1781). Type: Cape, Karoo, Thunberg in Herb. Thunberg 7796 (UPS, holo!).
Although C. capitella and C. spicata were described by Thunberg in the same publication the former name is prefered because it represents the more commonly found form of this species.
C. capitella is treated here as a species complex with five different subspecies which show an almost complete range from shrubs, to perennials with a basal rosette, to biennials. The petal apices vary from recurved to erect with the dorsal appendages becoming more enlarged and eventually being found in the terminal position.
1 Petal with smooth appendage in terminal position: 2 Bracts subtending the part-inflorescences glabrous; squamae emarginate . C. capitella subsp. capitella 2 Bracts subtending part-inflorescences densely hairy; squamae truncate . C. capitella.subsp. nodulosa I Petal with denticulate appendage in dorsal position: 3 Sepals covered with recurved hairs and marginal cilia; squamae oblong. Cc. capitella subsp. enantiophylla 3 Sepals glabrous or with a few marginal teeth; squamae transversely oblong to square: 4 Leaves with obtuse or rounded apices; petals scarcely recurved . F C. capitella subsp. meyeri 4 Leaves with acute apices; petals recurved to about half their length . C. capitella subsp. thyrsiflora
— subsp. enantiophylla (Bak. f.) Toelken, comb. nov. et stat. nov. C. enantiophylla Bak. f. in Bull. Herb. Boiss. ser. 2, 3: 816 (1903). Type: Trans- vaal, Houtbos, Rehmann 6372 (Z, holo!; BM!).
subsp. meyeri (Harv.) Toelken, comb. nov. et stat. nov. C. meyeri Harv., Fl. Cap. 2: 344 (1862). Type: Natal, between the Omsamculo and Omcomas Rivers, Drege s.n. (S, holo!; G!; P!).
subsp. nodulosa (Schonl.) Toelken, comb. nov. et stat. nov.
C. nodulosa Schonl. in Rec. Albany Mus. 1: 56 (1903) et in Trans. Roy. Soc. S. Afr. 17: 247 (1929). Type: Cape, Warrenton, Adams 28 (GRA, holo!; PRE!; SAM!).
— subsp. thyrsiflora (Thunb.) Toelken, comb. nov. et stat. nov.
C. thyrsiflora Thunb. in Acta Nova Phys.-Med. Acad. Caes. Leop.-Carol., Nat. Cur. 6: 329, 339 (1778) et Fl. Cap. ed. Schultes 283 (1821); L. f., Suppl. 190 (1781). Type: Cape, between Sundays and Fish River, Thunberg in Herb. Thunberg 7801 (UPS, holo!).
C. turrita Thunb. in Acta Nova Phys.-Med. Acad. Caes. Leop.-Carol., Nat. Cur. 6: 329, 336 (1778) et Fl. Cap. ed. Schultes 285 (1823); L. f., Suppl. 189 (1781). Type: no type preserved.
Vv
New Taxa and New Combinations in the Genus Crassula 101
13. Crassula congesta N.E. Br. subsp. laticephala (Schonl.) Toelken, comb. nov. et stat. nov.
C. laticephala Schonl. in Rec. Albany Mus. 2: 457 (1913). Type: Cape, Rivers- dale, Volschenk s.n. (GRA, holo!; K!; SAM!).
The subsp. congesta which occurs to the north of the Witteberge is disting- uished by its blunt leaf apices which are curved upwards, while the subsp. laticephala with its acute recurved leaf apices occurs to the south-east of these mountains.
14. Crassula corallina Thunb. subsp. macrorrhiza Toelken, subsp. nov. ab subsp. corallina radicibus principalibus tuberosis 0,3—0,8 (—1,2) cm in diametro, ramis sine radicibus adventiis et carpella stylis distinctis differt.
C. corallina sensu Higgins in J. S. Afr. Bot. 24: 115-118, fig. 2A (1958); sensu Friedr. in Prodr. Fl. S. W. Afr. 52: 26 (1968), non Thunb.
Plantae caulibus brevis erectis vel longioribus decumbentibus et sine radicibus adventiis vel rare solum in node singulare, radicibus principalibus tuberosis 0,3-0,8 (—1,2) cm in diametro. Folia angulato-obovata, fere obtriangularia, 0,4-0,5 cm longa, (0,35—) 0,4-0,5 cm lata, plerumque obtusa, texta strato cerae in paginis exposis. Carpella stylis distinctis ovariis recurvis ad apicem et stigmati- bus sessilibus lateralibus.
Type: South West Africa, Klein Karas, Dinter 4773 (PRE, holo!; K!).
Plants with short erect, or with longer decumbent or prostrate stems which are usually without adventitious roots, or, if present, then only at a single node at a time, with tuberous main roots 0,3—0,8 (-1,2) cm in diameter. Leaves angular-obovate to almost obtriangular, 0,4-0,5 cm long, (0,35—) 0,4-0,5 cm broad and usually with rounded apex, covered with a white waxy layer on the exposed surfaces. Carpe/s with indistinct styles and the broad upper part of the ovary recurved at the apex and with sessile lateral stigmas.
The subsp. macrorrhiza is restricted to Bushmanland and south-eastern South West Africa. SOUTH WEST AFRICA—2718 (Grunau): Klein Karas (-CA), Dinter 4773 (PRE, K). —2818 (Warmbad): Warmbad (-BC), Herre in SUG 5773 (GRA). CAPE—2717 (Vioolsdrif): Stinkfontein Plateaux, Pearson 6204 (GRA). —2818 (Warmbad): 20 km east-south-east of Goodhouse (-CD), To/ken 3681 (BOL). —2820 (Kakamas): Kakamas (-DC), Fuller 32 (BOL). —2821 (Upington): Upington (-AC), Simpson s.n. (NBG). —2917 (Springbok): 40 km north of O’Okiep (-BD), Horrocks 17 (NBG). —2918 (Gamoep): Aggenys (-BB), Pearson 2945 (BOL, GRA, SAM).
102 Journal of South African Botany
—2919 (Pofadder): near Pofadder (-AB), Barker 8357 (NBG). —2921 (Kenhardt): 4 km west of Kenhardt (-AC), Acocks 16 507 (PRE).
15. Crassula decumbens Thunb. var. brachyphylla (Adamson) Toelken, comb. nov. et stat. nov. C. brachyphylla Adamson in J. S. Afr. Bot. 8: 276 (1942). Type: Cape, Devil’s Peak, Ecklon & Zeyher 1852 (G!; GRA!; K!; S!; S!; SAM!).
The var. brachyphylla is usually distinguished from the typical variety by its club-shaped leaves and the broadly ovate petals, but plants from Bredasdorp are very similar to those of the var. decumbens.
16.Crassula elegans Schonl. & Bak. f. subsp. namibensis (Friedr.) Toelken, comb. noy. et stat. nov. C. namibensis Friedr. in Mitt. Bot. Staatssamml. Munchen 6: 632, fig. 5 (1967) et in Prodr. Fl. S. W. Afr. 52: 33 (1968). Type: South West Africa, Liideritz- bucht, Merxmiiller & Giess 3477 (M, neo).
The subsp. namibensis occurs in rock outcrops near Luderitz Bay and is distinguished from the typical subspecies by its blunt somewhat swollen tri- chomes on the leaves.
17. Crassula elsieae Toelken, sp. noy. ab C. corallina hydathodis marginalibus, squamis oblongo-cuneatis, habitione; ab C. papillosa foliis obovatis et sepalis - oblongis differt.
Herbae perennes ramis prostratis vel decumbentibus usque ad 10 cm longis et rare 0,1 cm in diametro latioribus et non lignosis ad basim, radicibus adventiis secus ramum, foliis veteribus non caducis. Folia sessilia, obtusa et saepe ali- quantum angulata, 0,2—-0,4 cm longa, 0,2-0,3 mm lata et basi cuneata plana sed succulentescentia ad apicem, minute papillosa saltem ubi juvenia, viridia vel fusca; vaginae foliorum circiter 0,05 mm longa et membranacea. Hydathodi aliquot plus minusve seriales in marginibus foliorum, vix visibiles. /nflorescentia dichasium terminale (1—) 3—5 floribus pedicellatis et pentameris, pedunculo indistincto et bracteis ellipticis succulentibus. Sepala late oblonga, 1,5-2 mm longa, obtusa, succulenta, viridia. Petala oblonga vel oblanceolata, 2,5-3 mm longa, obtusa sine appendicibus dorsalibus, connata in tubum 0,2-0,4 mm longem, loborum apicis recurvis, albis vel eburneis. Stamina circiter 2 mm longa, antheris luteis 0,2-0,3 mm longis, filamentis aliquantum latentibus ad basim et vix constrictis ubi connatis petalorum tubo. Squamae oblongo-cuneatae, 0,4-0,6 » 0,3-0,4 mm, truncatae et aliquantum constrictae ad basim, vix succulentae, pallide flavae. Carpella ovariis late reniformibus gradatim constrictis in stylos breves stigmatibus terminalibus indistinctis; ovarium laeve et 4-6 ovulis elongatis dense tectum papillis serialis saepe indistinctis.
_
V
New Taxa and New Combinations in the Genus Crassula 103
Type: Cape, Krakadouw Peak, Esterhuysen 14 958 (BOL, holo!).
Perennials with prostrate or decumbent branches up to 10 cm long and rarely more than | cm in diameter and not woody towards the base, with adventitious roots along the stem, with old leaves remaining attached to the stem. Leaves sessile, obovate and often somewhat angular, 0,2-0,4 cm long, 0,2-0,3 mm broad, obtuse and with a cuneate base, flat but becoming more fleshy towards the apex, minutely papillose at least when young, green to brown; sheath c. 0,05 mm long and membranous. Hydathodes few, more or less clearly arranged in a row on the margin of the leaves, hardly visible. Inflorescence a terminal dichasium with (—1) 3-5 flowers pedicellate and 5-merous, with peduncle hardly distinguishable from the vegetative branches and with fleshy elliptic bracts. Sepals broadly oblong, 1,5-2 mm long, with rounded apex, fleshy green. Petals oblong to oblanceolate, 2,5-3 mm long, obtuse and without dorsal appendage, fused into a tube 0,2-0,4 mm long with apices of the lobes recurved, white or cream. Stamens c. 2 mm long, with yellow anthers 0,2-0,3 mm long, with filaments slightly broadened towards the base and hardly constricted where fused to the petal tube. Squamae oblong-cuneate, 0,4-0,6 x 0,3-0,4 mm, truncate and somewhat constricted towards the base, slightly fleshy, pale yellow. Carpels with broadly reniform to almost pear-shaped ovaries gradually con- stricted into short styles with indistinct terminal stigmas; ovary smooth and with 4-6 elongate ovules densely covered with often indistinct rows of papillae.
In shaded places under rocks or in rock basins and known only from the northern Cedarberg.
This species is very rare and all the plants recorded of it were collected by the very observant collector Miss Elsie Esterhuysen, in honour of whom the species is named.
CAPE—3219 (Wuppertal): Krakadouw Peak (-AA), Esterhuysen 14 958 (BOL); Krakadouwsberg (-AA), Esterhuysen 7517 (BOL).
18. Crassula ericoides Haw. subsp. tortuosa Toelken, sp. nov. ab subsp. ericoides prostrato habitu et ovariis ovulis quatuor differt.
Plantae prostratae vel decumbentes, ramis precique ex base et ramis princi- palibus tortuosis. Folia (0,2—) 0,3-0,4 cm longa rare usque ad 0,1 cm lata et foliis veteribus remanentibus in ramis. Petala 0,2-0,3 cm longa, aliquantum expansa et saepe leviter recurva. Carpella circiter 0,2 cm longa et ovulis quatuor.
Type: Cape, top of Prince Alfred’s Pass, Té/ken 3757 (BOL, holo!).
Plants prostrate or decumbent, branching mainly from the base and with main branches usually tortuous. Leaves (0,2-) 0,3-0,4 cm long, rarely broader than 0,1 cm and old leaves remaining attached to the branches for some time. Petals 0,2-0,3 (-0,4) cm long, spreading and often slightly recurving. Carpels c. 0,2 cm long and with 4 ovules.
\//
104 Journal of South African Botany
On shallow soil on rocks at high altitude in the mountains from the Great Winterhoek to the Outeniqua mountains. CAPE—3321 (Ladismith): Cloetes Pass (-DD), Zinn in SAM 59 071.
—3322 (Oudtshoorn): Spitskop (-BC), Thorne in SAM 5/ 442.
—3323 (Willowmore): top of Prince Alfred Pass (-CC), T6/ken 3757 (BOL); near Joubertina (-DD), Esterhuysen 13 648 (BOL).
—3324 (Steytlerville): Winterhoek mountains, Zeyher 2523 (PRE, SAM).
19. Crassula exilis Harv. subsp. exilis C. exilis Harv., Fl. Cap. 2: 347 (1862); Schonl. in Trans. Roy. Soc. S. Afr. 17: 252 (1929). Type: Cape, Namaqualand, Whitehead s.n. (TCD, holo!).
C. exilis has often been wrongly interpreted as its typical form has very fleshy leaves and a clearly defined peduncle, unusual features within the section Rosulares, but the variation within the species complex shows a complete range of these characters.
1 Peduncle with short spreading blunt hairs (usually not visible without lens) .
C. exilis subsp. exilis I Peduncle glabrous. : C. exilis subsp. sedifolia I’ Peduncle with spreading slightly recurved hairs c. 11 mm Jong . . C. exilis subsp. cooperi
— subsp. sedifolia (N. E. Br.) Toelken, comb. nov. et stat. nov.
C. sedifolia N. E. Br. in Gard. Chron. ser. 3, 32: 429 (1902); Schonl. in Trans. Roy. Soc. S. Afr. 17: 236 (1929). Type: South Africa without precise locality, MacOwan s.n. (K, holo!).
—— subsp. cooperi (Regel) Toelken, comb. nov. et stat. nov. C. cooperi Regel in Gartenflora 36, t. 786 (1874); Schonl. in Trans. Roy. Soc. S. Afr. 17: 236 (1929). Type: South Africa, without precise locality. (LE, holo!).
20. Crassula expansa Dryand. in Ait. subsp. expansa
C. expansa Dryand. in Ait., Hort. Kew. ed. 1, 1: 390 (1789); Harv., Fl. Cap. 2: 354 (1862); Schonl. in Trans. Roy. Soc. S. Afr. 17: 184 (1929). Type: Caput Bonae Spei, Masson s.n. (BM, holo!).
This very variable species complex of which five different subspecies occur in Southern Africa is very widespread in Africa. Some of the subspecies often occur geographically very close to one another but usually remain ecologically isolated.
1 Sepals and leaves covered with spreading hairs: 2 Hydathodes on margin of leaves; moist places on Swartberg
C. expansa subsp. peculiaris 2 Hydathodes scattered over the upper leaf surface; northern Natal and Transvaal C. expansa subsp. fragilis
New Taxa and New Combinations in the Genus Crassula 105
I Sepals and leaves glabrous: 3 Leaves with one or two rows of hydathodes; stems without adventitious roots: widespread from northern Natal to southern and northern Cape eee eee C. expansa subsp. expansa 3 Leaves usually with two rows of hydathodes; stems with woody adventitious roots; along the coast from Port Elizabeth to the Cape Peninsula . Fen Lk chile C. expansa subsp. filicaulis 3’ Leaves with hydathodes scattered at least over the upper surface; stems usually with adventitious roots; Namaqualand and into south-western South West Africa C. expansa subsp. pyrifolia
—— subsp. filicaulis (Haw.) Toelken, comb. nov. et stat. nov. C. filicaulis Haw. in Phil. Mag. 24: 187 (1824); DC., Prodr. 3: 384 (1828); Eckl. & Zeyh., Enum. 295 (1837), partly. Iconotype: plate 768 (K, lecto!).
—— subsp. fragilis (Bak.) Toelken, comb. nov. et stat. nov.
C. fragilis Bak. in J. Linn. Soc. (Bot.) 22: 469 (1887). Type: central Madagascar, Baron 3348 (K, holo!).
C. browniana Burtt Davy, Fl. Transv. 141 (1925); Schonl. in Trans. Roy. Soc. S. Afr. 17: 185 (1929). Type: Transvaal, near Lydenburg, Wilms 518 (K, holo!).
—— subsp. peculiaris Toelken, subsp. nov. ab subspeciebus aliis foliis hirsutis et radicibus non gralliformibus; ab subsp. fragile floribus axillaribus et habitate altitudine differt.
Herbae perennes ramis prostratis usque ad 30 cm longes et interdum radicis adventiis gracilibus. Folia petiolis usque ad 0,3 cm longis; lamina ovata vel late elliptica (0,2—) 0,4-0,6 cm lata, dorsoventraliter complanata, tecta pilis gracilibus et hydathodis paucis secus margines. Flores singulares in axilis foliorum et saepe non ad apices ramorum. Sepa/a longitudine c. 3/4 petalis partes aequantia, acuta vel obtusa et pubescentia.
Type: Cape, east of Oliewenberg, Esterhuysen 28 832 (BOL, holo!).
Perennials with prostrate herbaceous branches up to 30 cm long, forming irregular mats and sometimes with fine adventitious roots. Leaves with petioles up to 0,3 cm long; lamina ovate to broadly elliptic, (0,2—) 0,4-0,6 cm broad, dorsiventrally compressed, covered with long fine hairs and with few hydathodes along the margin. Flowers borne singly in the axils of the leaves and often not even at the apical part of the stem. Sepals three-quaters of the length of the petals, bluntly acute and covered with scattered hairs.
On moist shaded areas and usually under rocks at an altitude of above 1 500 m and recorded only from the vicinity of the Swartberg Pass.
CAPE—3322 (Oudtshoorn): Swartberg Pass (-AC), Esterhuysen 4559 (BOL); east of Oliewenberg (-AC), Esterhuysen 28 832 (BOL).
106 Journal of South African Botany
— subsp. pyrifolia (Compton) Toelken, comb. nov. et stat. nov. C. pyrifolia Compton in J. Bot. Lond. 70: 283 (1932). Type: Cape, near Bitter- fontein, Moller sub Compton 3954 (BOL, holo!; K!).
21. Crassula garibina Marl. & Schonl. subsp. glabra Toelken, subsp. nov. ab subsp. garibina foliis glabris vel rare papillis minutis sine apicibus sphaericis differt.
Plantae praeter calycem minute papillosam glabrae. Folia viridia vel viridi-flava et glabra rare juvenilia minute papillosa.
Type: Cape, Geselskapbank, Té/ken 3687 (BOL, holo!).
Plants glabrous except for minutely papillose calyx. Leaves green to yellowish- green and glabrous or rarely minutely papillose when young. CAPE—2917 (Springbok): Eenriet mountain (-BB), Smithers s.n. (BOL); 10 km north of Steinkopf (-BB), To/ken 4324 (BOL).
—2918 (Gamoep): Geselskapbank (-AA), Télken 3687 (BOL).
22. Crassula globularioides Britt. subsp. argyrophylla (Schonl. & Bak. f.) Toelken, comb. nov. et stat. nov.
C. argyrophylla Diels ex Schonl. & Bak. f. in J. Bot., Lond. 40: 290 (1902); Diels in Bot. Jahrb. 39: 465 (1907); Burtt Davy, Fl. Transv. 141 (1926). Syntypes: Transvaal, Braamfontein, Gilfillan sub Galpin 62/1] (GRA!; BOL!; PRE!); Lydenburg, Wilms 527 (BM!)
subsp. illichiana (Engl.) Toelken, comb. nov. et stat. nov.
C. illichiana Engl. in Diels et Engl. in Bot. Jahrb. 39: 465 (1907). Syntypes:
Tanzania, Usambara, Buchwald 174 (K!); Manca near Sakara, Engler 1073 (B). The subsp. argyrophylla is distinguished from the very similar subsp.
illichiana by only one bract without part-inflorescences on the peduncle, and
its marginal cilia becoming shorter towards the base of the leaves.
23. Crassula grammanthoides (Schonl.) Toelken, comb. nov. et stat. nov. Dinacria grammanthoides Schonl. in Bull. Herb. Boiss. ser. 1, 5: 859 (1897). Type: Cape, Swartberg near Caledon, Schlechter 5570 (GRA, holo!; BOL!).
C. grammanthoides is very similar to C. depressa except that instead of a terminally placed stigma it is here in a lateral position. This represents an intermediate stage between species of Crassula, such as C. depressa, and Dinacria filiformis and Dinacria sebaeoides where the styles are variously enlarged to bring the stigma into a lateral position.
24. Crassula lanuginosa Hary. var. pachystemon (Schonl. & Bak. f.) Toelken, comb. nov. et stat. nov.
New Taxa and New Combinations in the Genus Crassula 107
la cm
"@ pachystemon Schonl. & Bak. f. in J. Bot., Lond., 36: 367 (1898); Schonl. in Trans. Roy. Soc. S. Afr. 17: 258 (1929). Type: Cape, near Graaff Reinet, Bolus 437 (GRA, lecto!; K!; PRE!).
C. ernestii Schonl. & Bak. f. in J. Bot., Lond. 40: 283 (1902); Schonl. in Trans. Roy. Soc. S. Afr. 17: 258 (1929). Type: Cape, Bowker’s Park, Galpin 2563 (GRA, holo!; BM!).
The var. /anuginosa is distinguished from the typical variety by its usually smaller leaves with an acute apex and one to few terminal hairs. The two varieties occur without any intermediates in the vicinity of Cradock but between Tarkastad and Queenstown some intermediates have been found.
25. Crassula latibracteata Toelken, sp. nov. ab C. cultrata et C. rogersio petalis oblongo-oblanceolatis et thyrso capitato circumcinato bracteis late oblanceolatis differt.
’.Frutices perennes usque ad 40 cm alti, plerumque multum ramosi et ramis superis rectis et erectis, infernis saepe decumbentibus 0,1—0,3 cm in diametro et cortice chartaceo, foliis veteribus deciduis. Folia sessilia, obovata vel ali- quantum falcata, (0,8—) 1-2 (—2,5) cm longa, 0,4-0,7 cm lata, obtusa dorso- ventraliter complata et leviter convexa sed marginibus argutis, puberula vel glabra, viridia ad lutea-viridia et saepe marginibus corneis rubris; vaginae foliorum 0,1-0,2 cm longae. Hydathodi diespersae in paginis superibus et infernis, conspicuae. /nflorescentia thyrsus terminalis densus et fere capitatus 1—2 dichasiis floribus pentameris sessilibus, pedunculo puberulo vel glabrescente 15-25 cm longo, bracteis obovatis ad late oblongis saepe marginibus mem- braneis et obtangentibus inflorescentiis. Sepala oblongo-triangularia, 2-2,5 mm longa, obtusa, succulentissima ad apicem, puberula, viridia vel luteo- viridia. Petala oblonga vel oblongo-oblanceolata, 3—3,5 mm longa, appendicibus dorsalibus elongatis, 0,5-0,8 mm connata, eburnea. Stamina 2,5-3 mm longa, antheris luteis 0,5-0,8 mm longis, filamentis non latioribus ad basim et ali- quantum constrictis ubi connatis tubo petalorum. Squamae oblongo-cuneatae, 0,6-0,8 x 0,4-0,6 mm, truncatae vel leviter emarginatae, plerumque aliquantum constrictae ad basim, leviter succulentae, luteae. Carpella ovariis reniformibus abrupte constrictis ad styles breves crassos stigmatibus latis viridibus et saepe aliquantum lateralibus; ovarium ciliis brevibus secus suturam et 10-14 ovulis elongatis porcis leviter papillosis.
Type: Cape, Riebeeck East, Télken 4344 (BOL, holo!).
Perennials forming shrublets up to 40 cm high, usually much branched and with upper branches erect and straight, with lower branches often wiry and decumbent 0,1-0,3 cm in diameter and with peeling bark, with old leaves deciduous. Leaves sessile, obovate to somewhat falcate and turned to the one
108 Journal of South African Botany
side of the stem, (0,8—) 1-2 (-2,5) cm long, 0,4-0,7 cm broad, with rounded apex, dorsiventrally flattened and slightly convex on either surface but with sharp margins, puberulous to glabrous, green to yellowish-green and often with red horny margin; sheath 0,1-0,2 cm long and formed by the partial fusion of the leaf bases. Hydathodes scattered over the upper and lower leaf margin, more or less conspicuous. Inflorescence a terminal dense almost capitate thyrse with one to three dichasia with sessile 5-merous flowers, with peduncle puberulous to glabrascent 15-25 cm long with obovate to broadly oblong bracts often with a membranous margin covering the lower parts of the inflo- rescence. Sepals oblong-triangular, 2—-2,5 mm long, with rounded apices often very swollen so that it appears to be recurved, puberulous, green to yellowish- green. Petals oblong to oblong-lanceolate, 3-3,5 mm long, each with elongate dorsal appendage in terminal position, fused into a tube 0,5—0,8 mm long, with short swollen apices curved downwards, cream. Stamens 2,5-3 mm long, with yellow anthers 0,5—0,8 mm long, with filaments not broadened towards the base and somewhat constricted where fused to the petal tube. Squamae oblong- cuneate, 0,6-0,8 x 0,4-0,6 mm, truncate or slightly emarginate, usually some- what constricted towards the base, slightly fleshy, yellow. Carpels with reniform ovaries abruptly constricted into short stout styles ending in broad green and somewhat laterally placed stigmas; ovary with short cilia along the sutureand with 10-14 elongate ovules covered with faintly papillose ridges. CAPE—3226 (Fort Beaufort): Adelaide (-BC), Rogers 27 680 (PRE).
—3325 (Port Elizabeth): Suurberg (-BD), Barker 4941 (NBG).
—3326 (Grahamstown): Riebeeck East (-AA), Tolken 4344 (BOL); near Southwell (-CA), Acocks 11 086 (PRE).
26. Crassula mesembryanthoides (Haw.) Dietr. subsp. hispida (Haw.) Toelken, comb. nov. et stat. nov. Globulea hispida Haw. in Phil. Mag. 66: 30 (1825). Iconotype: plate 762 (K, lecto!). Crassula hispida (Haw.) Dietr., Syn. Pl. 2: 1031 (1840).
The internodes are very short (0,3-0,5 cm long) at the base of the branches in the subsp. hispida while they are usually 1-1,5 cm long in the subsp. mesem- bryanthoides. A a) 7)
Mi al lt }
27. Crassula minuta Toelken, sp. nov. ab C. vaillantio sepalis petales aequan- tibus et squamis transverse oblongis et abrupte cuneatis; ab C. decumbenti foliis ad 3 mm longes et floribus terminalibus differt.
Herbae annuae ramis erectis 1-3 cm longis leviter rigidis, glabrae, internodiis aequalibus. Folia sessilia, linearia vel elliptica, 0,1—-0,2 (-0,3) cm longa, 0,05-0,1 cm lata, plerumque obtusa, plus minusve plana supra et convexa subtus, glabra, succulenta, viridia vel testacea; vaginae foliorum c. 0,05 cm longa, membranacea.
2
ie
New Taxa and New Combinations in the Genus Crassula 109
Hydathodi paucae secus marginem, inconspicuae. Inflorescentia plerumque flores singulares terminales et pedicellis brevibus. Sepala elliptico-lanceolata, 1—1,5 mm longa, acuta, glabra, viridia vel testacea. Peta/a lanceolata, 1-1,5 mm longa, obtusa et sine appendicibus dorsalibus, connata tubo ad 0,3 mm longam, effusa alba vel eburnea et saepe rubrotincta. Stamina c. 1 mm longa, antheris flavis c. 0,1 mm longis et latis, filamentis vix latis factis ad bases. Squamae transverse oblongatae et abrupte cuneatae, 0,4-0,5 x 0,5-0,6 mm, truncatae abrupte constrictae ad bases, paene membranaceae, pallide flavae vel albae. Carpella ovariis oblongo- reniformibus gradate constrictis ad styles breves et stigma- tibus terminalibus; ovarium leve et 2-4 ovulis elongatis leviter papillosis.
t
Type: Cape, Scorpionsberg, Esterhuysen 12 233 (BOL, holo!; PRE!).
Annuals with erect branches 1-3 cm long and usually more or less wiry, glab- rous, with pairs of leaves at about equal distances from one another. Leaves sessile, linear-elliptic, 0,1-0,2 (—0,3) mm long, 0,05-0,1 cm broad, usually obtuse, more or less flat above and convex below, glabrous, fleshy, green to reddish-brown; sheath c. 0,05 cm long, membranous. Hydathodes few along the margin, inconspicuous. Jnflorescence usually single 4-merous flowers on short pedicels at the end of branches. Sepals elliptic- lanceolate, 1—1,5 mm long, bluntly, acute, glabrous, green to reddish-brown. Petals lanceolate, 1-1,5 mm long, obtuse and without dorsal appendages, fused into a tube 0,3 mm long, spreading, white or cream and often tinged red. Stamens c. 1 mm long and broad, with filaments scarcely broadened downwards and hardly fused to the tube. Squamae transverely oblong and abruptly cuneate, 0,4-0,5 = 0,5—0,6 mm, truncate, abruptly constricted towards the base, almost membranous, pale yellow or white. Carpels with oblong-reniform ovaries gradually constricted into short styles and with terminal stigmas; ovary smooth and with 2-4 elongate ovules which are faintly papillose. = CAPE—3219 (Wuppertal): Scorpionsberg (-AC), Esterhuysen 12 233 (BOL, PRE); Tafelberg (-AC), Esterhuysen 13 854 (BOL). 28. Crassula montana Thunb. subsp. quandrangularis (Schonl.) Toelken, comb. nov. et stat. nov. C. quadrangularis Schonl. in Rec. Albany Mus. 1: 57 (1903). Type: Cape, Laingsburg, Marloth 2512 (GRA, holo!).
The squamae are scarcely constricted towards the base and the inflorescence is a flat-topped thyrse in the subsp. quadrangularis.
29. Crassula multicava Lem. subsp. floribunda Friedr. ex Toelken, subsp. nov. ab subsp. multicava inflorescentia floribus pentameris et sine gemmis adyentitis differt.
110 Journal of South African Botany
Plantae erectae ramis rigidis 0,3-0,6 cm in diametro ubi foliis et glabrae vel hirsutae. Inflorescentia dichasiis floribus multis, braceis liniaribus apicibus obtusis, sine gemmis adventitis et floribus pentameris. Petala 5-6 mm longa et remanentia post florescentiam, alba vel eburnea. Ovula elongata et circiter dupla longiora suo diametro, tecta seriebus papillarum acutarum.
Type: Natal, old Tugela bridge (-AB), Strey 4209 (PRE, holo!; NH!).
Plants erect with stiff branches 0,3-0,6 cm in diameter when with leaves, and the whole plant glabrous to hairy. Inflorescence with many-flowered dichasia with narrowly oblong bracts with blunt apices, without adventitious buds, with 5-merous flowers. Petals 5-6 mm long, recurved and remain so after flowering, white or off-white. Ovules elongate to about twice as long as broad, covered with rows of pointed papillae.
NATAL—2931 (Stanger): old Tugela bridge (-AB), Strey 4209 (NH, PRE). 2930 (Pietermaritzburg): Inanda (-DB), Wood 597 (BOL, SAM). Durban Botanic Gardens, Wood 11 744 (NH, PRE, SAM).
30. Crassula multiflora Schonl. & Bak.f. subsp. leucantha (Schonl. & Bak.f.) Toelken, comb. nov. et stat. nov. C. leucantha Schonl. & Bak.f. in J. Bot., Lond. 36: 369 (1898); Schonl. in Trans. Roy. Soc. S. Afr. 17: 262 (1929). Type: Cape, Houwhoek, Schlechter 7378 (GRA, holo!; BM!; BOL!; G!; K!; P!; Z!).
This plant is known only from the type collection which differs from plants of typical C. multiflora by its oblanceolate leaves which are rarely longer than 2,5 cm.
31. Crassula muscosa L. var. muscosa
C. muscosa L., Pl. Rar. Afr. 10 (1760) et Mantissa altera 361 (1771); Harv., Fl. Cap. 2: 351 (1862), partly, excl. specimens cited; Tolken in J. S. Afr. Bot. 38: 70 (1972). Type: Cape, sine leg. in Herb. Burmann (G, holo!).
C. lycopodiodes Lam., Dict. 2: 173 (1786); Harv., Fl. Cap. 2: 351 (1862); Schonl. in Trans. Roy. Soc. S. Afr. 17: 189 (1929); Friedr. in Prodr. Fl. S. W. Afr. 52: 31 (1968). Type: Africa, sine leg. in Herb. Lamarck (P—LA, holo!).
The different varieties of this species complex often grow in different ecolo- gical niches but equally often they are found next to one another.
C. polpodacea and C. propinqua which seem to refer to the var. sinuata and var. rigida respectively were not adopted, because their diagnoses are so vague and the original material is often confused so that the taxa concerned cannot be recognized without doubt. C. parvula is, however, easily interpreted.
New Taxa and New Combinations in the Genus Crassula 111
1 Stems at base without old leaves and with peeling bark; leaves yellow-green and up to 0,1 cm long C. muscosa var. rigida
1 Stems with old leaves at the ‘base: leaves grey- green 0, 2-0 4 (- 0 6) cm long: Squamae emarginate; stems wiry-woody with ‘leaves adpressed 0,2-0,3 cm in
diameter and up to 40 cm long . C. muscosa var. sinuata
2 Squamae truncate; stems woody or ‘if wiry then with spreading leaves and erect habitat:
3 Stems with leaves 0,3-0,4 (-0,6) cm in diameter; plants more or less
branched and up to 60 cm long . . C. muscosa var. muscosa
3 Stems with leaves c. 0,25 cm in diameter; plants erect tufts up to 15 cm
high . . kan ee . . .C. muscosa var. parvula
var. parvula (Eckl. & Zeyh.) Toelken, comb. nov. et stat. nov.
Tetraphyle parvula Eckl. & Zeyh., Enum. 294 (1837). Type: Cape, Botha’s Berg, Ecklon & Zeyher 1871 (S, lecto!).
Crassula parvula (Eckl. & Zeyh.) Endl. ex Walp., Repert. 2: 253 (1843); Harv., Fl. Cap. 2: 352 (1862).
—— var. rigida Toelken, var. nov. ab varietatibus aliis foliis usque ad 2 mm longis et obtusis, cortice cadenti membranaceo differt.
Plantae erectae 10-15 (-30) cm altae ramis principalibus ligneis et rigidis 0,2-0,5 (0,8) cm in diametro sine foliis sed contice candenti membranaceo, ramis multis ad basim, ramis juvenibus foliis usque ad 0,3 mm in diametro et internodiis visilibus inter folia praecipue in exsiccata sed foliis veteribus deciduis. Folia ovata vel late ovata, 0,05-0,1 (—0,2) cm longa et lata, obtusa, apice effuso et gemma axillare in quoque axilla. Squamae truncatae.
Type: Cape, Stinkfontein, Schlieben 11 530 (PRE, holo!; STE!).
Plants erect 10-16 (—30) cm high with rigid woody branches 0,2-0,5 cm thick without leaves and with peeling bark, branching mainly from the base and much branched, with younger branches with leaves up to 0,3 cm in diameter and with internodes visible between the leaves particularly in dried specimens, with old leaves deciduous. Leaves ovate to broadly ovate, 0,05—0,1 (—0,2) cm long and broad, obtuse, rarely more or less adpressed but with an axillary bud in the axil of all leaves, yellowish-green. Squamae truncate.
Occurs from the Cedarberg northwards through the mountains in Nama- qualand and Richtersveld to just north of the Orange River.
— var. sinuata Toelken, var. nov. ab varietatibus aliis ramis gracilibus sinuatis et squamis emarginatis.
Plantae ramis gracilibus sinuatis vel decumbentibus 30-50 (—80) cm longis, 0,2-0,3 (-0,4) cm in diametro foliis, minute ramosis et sine ramis axillaribus brevis, tenacibus et ligneis et internodiis non visibilibus inter folia adpressa, foliis veteribus non deciduis. Folia lanceolata 0,2-0,3 cm longa, 0,1—0,2 cm lata, acuta. Squamae emarginatae.
12 Journal of South African Botany Type: Cape, near Komga, Flanagan 1795 (BOL, holo!; GRA!; PRE!).
Plants with slender sinuous or decumbent branches 30-50 (—80) cm long, 0,2-0,3 (0,4) cm in diameter with leaves, little branched and without short axillary branches, tough and woody and with internodes not visible between the leaves, with old leaves remaining attached to the stem. Leaves lanceolate 0,2-0,3 cm long, 0,1-0,2 cm broad, acute, yellowish-green rarely pale green. Squamae distinctly emarginate.
Occurs mainly in the triangle between Willowmore, Port Elizabeth and King William’s Town.
32. Crassula namaquensis Schonl. & Bak.f. subsp. comptonii (Hutch. & Pillans) Toelken, comb. nov. et stat. nov. C. comptonii Hutch. et Pillans, Botanist in S. Afr. 149 & illustr. (1946). Type: Cape, Vanrhyn’s Pass, Hutchinson 765 (K, holo!). |
Within the species complex C. namaquensis the tendency for the rostrate petal apex to elongate from up to | mm in the subsp. namaquensis to 2 mm in subsp. comptonii and 2-3 mm in the subsp. /utea cannot be ignored. Similarly the petal colour changes from white or cream to pale yellow and bright yellow. The range of variation of the leaves found in the subsp. namaquensis covers the variation found in the whole species.
—— subsp. lutea (Schonl.) Toelken, comb. nov. et stat. nov.
C. namaquensis var. lutea Schonl. in Rec. Albany Mus. 1: 117 (1904) et in Trans. Roy. Soc. S. Afr. 17: 269 (1929). Type: Cape, Bokkeveld Karoo, Marloth 3238 (GRA, holo!; K!).
C. lutea (Schonl.) Friedr. in Mitt. Bot. Staatssamml. 11: 333 (1974).
33. Crassula natans Thunb. var. filiformis (Eckl. & Zeyh.) Toelken, comb. nov. et stat. nov.
Bullardia filiformis Eckl. & Zeyh., Enum. 290 (1837). Type: Cape, Platteklip, Ecklon & Zeyher 1850 (GRA!; K!; S!; SAM!).
Helophytum natans (Thunb.) Eckl. & Zeyh. var filiformis (Eckl. & Zeyh.) Harv., Fl. Cap. 2: 329 (1862), partly.
Crassula natans Thunb. forma filiformis (Eckl. & Zeyh.) Schonl. in Ann. Bolus Herb. 2: 49 (1917).
The var. filiformis is much finer than the var. natans and the flowers are rarely up to 1 mm long and its petals do not recurve. The former grows usually on moist or marshy soil on the edge of the water, while the latter is often found in pools and under those conditions develops the typical broad floating leaves.
PP
New Taxa and New Combinations in the Genus Crassula 113
34. Crassula nudicaulis L. var. herrei (Friedr.) Toelken, comb. nov. et stat. nov. C. herrei Friedr. in Mitt. Bot. Staatssamml. Miinchen 6: 640, fig. 10 (1967). Type: Cape, Holgat River, Herre in SUG 8126 (M, holo!).
The typical var. herrei with almost terete leaves, as illustrated by Friedrich (1967) is an extreme form of a plant that has less fleshy leaves and marginal cilia near Komaggas. This form is, however, distinguished from plants of the typical variety on the Kamiesberg by its flowers in the globular part-inflores- cences which are all directed upwards and not spreading in all directions.
—— var. platyphylla (Harv.) Toelken, comb. nov. et stat. nov. C. platyphylla Harv. in Fl. Cap. 2: 363 (1862). Type: South Africa, Drége 6986 (S, holo!; G!).
The var. platyphylla is distinguished from the var. nudicaulis by its short broadly oblong leaves with subdistant marginal cilia and it has usually more than 24 ovules per carpel.
35. Crassula obovata Haw. var. dregeana (Harv.) Toelken, comb. nov. et stat. nov.
, C. dregeana Harv., Fl. Cap. 2: 346 (1862); Schonl. in Trans. Roy. Soc. S. Afr.
17: 223 (1929). Type: Natal, between Omsamculo and Omcomas, Drege s.n. (S, lecto!; BM!; G!; K!; LU!; P!; SAM!; TCD!).
The leaves of plants of this variety are almost terete and their recurved marginal cilia are adpressed so that they are indistinguishable from the other hairs that cover the leaves.
36. Crassula pageae Toelken, nom. nov.
Pagella archeri Schonl. in Ann. Bolus Herb. 3: 67, t. 3A (921): Berger in Pflanzenfam. ed. 2, 18a: 400 (1930). Type: Cape, near Montagu Bath, Page in BOL 16 645 (GRA, holo!; BOL!; PRE!).
The genus Pagella was distinguished from Crassula by its fused carpels which are half sunk into the swollen stem, but a similar phenomenon, although less far-reaching, can be observed in for instance C. hirsuta and C. glomerata. C. umbellata and C. dodii which often grow together with C. pageae produce stunted forms where the internodes also do not elongate. Pagella archeri is thus included in the genus Crassula but the name cannot be transferred because of C. archeri Compton.
37. Crassula pellucida L. subsp. pellucida
C. pellucida L., Sp. Pl. ed. 1: 283 (1753); DC., Prodr. 3: 386 (1828); Schonl. in Trans. Roy. Soc. S. Afr. 17: 193 (1929); Télken in J. S. Afr. Bot. 38: 71 (1972). Iconotype: Dillenius, Hort. Elth. t. 100, fig. 119.
114 Journal of South African Botany
The subspecies of this species complex are all geographically isolated except the subsp. alsinoides which occurs in forested areas, while the other subspecies are often found in the more open vegetation nearby.
1 Leaf sheath glabrous: 2 Leaves ovate and abruptly constricted at the base and more or less fused to one another. . . . C. pellucida subsp. marginalis 2 Leaves elliptic to elliptic- lanceolate with a cuneate to subpetiolate base: 3 Receptacle and carpels remaining soft and membranous when fruiting; perennials with branches often longer than 20 cm long; south-western Cape as far north as the Winterhoekberge _ C. pellucida subsp. pellucida